This PDF file contains the front matter associated with SPIE Proceedings Volume 13018, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Smart transportation is of great significance to reduce regional carbon emissions and achieve the dual carbon goal. However, the current research does not deeply explore the long-term dynamic impact of intelligent transportation on transportation carbon emissions. Therefore, this paper selects 30 provinces in China, uses their fuel use data from 2005 to 2020 to calculate transportation carbon emissions, and divides different regions according to geographical location and national development strategy. The dynamic impact of intelligent transportation on transportation carbon emissions is analyzed based on the PVAR model. The results show that the size effect of the 30 provinces is much larger than the structure effect and the technology effect.The main impact indicators of transport carbon emissions are different in the different regions like Beijing-Tianjin-Hebei Urban Agglomeration, the Yangtze River basin and the central and western region.Based on the above research, this paper aims to provide policy recommendations on transportation carbon emission reduction for cities in different regions from the perspective of smart cities.

With the burst-up of transportation demand, the scale of urban rail transit expands, and the complexity of transit network enhances. Modern communication technology renders methods to collect passenger flow data with ease and accuracy. From traffic OD and travel time, there is no intuitive and quantitative method to collect and grasp the travel path and distribution of passenger flow. Modeling the rail transit network, establishing a quantitative impedance model from time, money and other comprehensive factors, simulating the travel path selection behavior of passengers in a quantitative way, and then transforming OD data into the distribution of passenger flow in the network, can understand the passenger flow operation rules, and also can better classify ticket business, which provides reference for business analysis of various operation investors, and has practical significance. Based on the current situation of Suzhou rail transit and OD passenger flow data, this paper establishes the rail transit network topology model, comprehensively establishes the passenger travel cost model from multiple perspectives, searches for paths, improves the effective path set in combination with the actual situation, and obtains the passenger flow distribution results of the simulated line network through multiple iterations, which helps simulation on the real-time passenger flow status.

Analyzing the interaction patterns between vehicles is an effective means of analyzing the causes of congestion and accidents on expressways. To analyze vehicle behavior from a micro perspective, this study was conducted using a driving simulator to investigate the vehicle operational behavior generated by participants on a specific road section. Through separate interviews, the study investigated the intentions and reasons behind the participants' behavior during the replay process. The study decomposed and extracted 24 key words and structured combination the interview responses into four operational intention groups. The study found that the vehicle behavior corresponding to each operational intention group was significantly different. The analysis of the participants' operational behavior in different driving environments based on the relationships between operational intentions can provide policy recommendations for relevant transportation departments and provide theoretical support for driving behavior analysis.

With the increasing popularity of the idea of giving way to pedestrians in Chinese cities, drivers need to frequently perform eye movements such as gaze and scan during the driving process to avoid accidents and give way to pedestrians in a timely manner. They also need to obtain road information through traffic facilities on the road. This article uses UC-win/road software to build a 3D simulation road from the perspective of driver's visual characteristics and vehicle operating characteristics. The data is obtained through eye tracker and driving simulator technology, and a scenario of traffic facilities combination based on two different deceleration markings is set up. Analysis shows that the use of fishbone-shaped deceleration markings has the best effect on improving road safety and reducing the risk of accidents.

Reasonable transit transfer discount policies incentivize people to take public transportation and alleviate the increase in vehicle usage. This research establishes and analyzes a dynamic game model with perfect information for public transportation operator and travelers as players. By treating the optimization process of discount as a learning process for operator to continuously explore and obtain the optimal price in a complex traveler environment, an adaptive discount adjustment method based on deep reinforcement learning is proposed. The result shows that transfer discount can increase the travel share of public transportation; Operator can use the first mover advantage to and achieve the most favorable Nash equilibrium for himself; The applicability of the proposed method in transfer discount scenarios is tested via simulation; According to the reinforcement learning DDPG algorithm, the Nash equilibrium is reached when the fare change interval is between (-0.68, -0.52) yuan, resulting in an increase in the usage of public transportation.

China's dual-point policy has been instrumental in promoting the new energy vehicle industry. However, the adoption of battery swapping stations in the battery-as-a-service (BaaS) model poses significant cost and implementation challenges. To address this, the government has extended its support to the BaaS model. This study employs an evolutionary game model involving new energy enterprises and local governments to assess the impact of the dual-point policy and BaaS model subsidy policy, while also considering the government's environmental concerns, including carbon emissions. The key findings are as follows: Firstly, solely implementing the dual-point policy proves insufficient for widespread BaaS model adoption. Secondly, exclusively implementing the subsidy policy for swapping stations may lead to subsidy fraud. Lastly, the combined effect of the dual-point policy and the BaaS model subsidy policy is positive. Lower subsidy rates and credit prices attract vehicle enterprises towards the BaaS model.

The bus stop area is an important area for vehicles to generate traffic emissions. In order to study the characteristics of vehicle emissions in the bus stop area under the net-connected environment, a mixed traffic flow simulation environment is constructed based on the analysis of the travel behavior of manual driving vehicles, ACC vehicles and CACC vehicles, and the influence of different factors on vehicle carbon emissions in the bus stop area is analyzed by using the car-following model and lane-changing model. Firstly, the full speed difference model, ACC model and CACC model are selected as the car-following models for manual driving vehicles and net-connected vehicles respectively, and the lane-changing model is constructed using the discretionary lane change model and the mandatorylane change model; the VT-Micro model is selected as the carbon emission calculation model; finally, MATLAB is used for numerical simulation to simulate and analyze the traffic emission distribution characteristics of the bus stop area under the net-connected environment. The results show that the carbon emissions in the upstream of the bus stop increase as the proportion of bus increases, the traffic volume increases and the proportion of net-connected vehicle decreases. The carbon emission hotspot was identified as the upstream of the stop near the inner lane. This result can provide a basis for carbon emission management at bus stops in the future.

With the continuous development of urban transportation planning, how to accurately obtain information on individual travel patterns has become the core problem of research. However, traditional methods of residential travel surveys, such as paper questionnaires, telephone interviews, and mail inquiries, are limited by low data accuracy, organizational difficulties, and limited sample size, which cannot meet the needs of modern transportation planning. GPS-based travel survey methods have gradually become the focus of researchers' attention due to their advantages of high accuracy positioning, convenience, and no human intervention. In this paper, we propose a new method for individual travel mode recognition using cell phone GPS positioning data. Firstly, through a multi-mode travel trajectory data collection application based on smartphone GPS sensors, the characterization indexes of different travel modes are extracted. Secondly, this paper proposes a temporal convolutional neural network algorithm for travel mode recognition, which has strong stability and self-adaptability, and combined with smoothing optimization can effectively improve the accuracy rate of mode recognition. The experimental results show that the average accuracy of mode detection for walking, bicycle, bus and car reaches more than 96%, which verifies the feasibility and effectiveness of the method in this paper. The research results of this paper can provide strong support for future urban transportation planning and travel behavior research.

Natural wind and the road itself brings the airflow changes, together constitute the vehicle driving process of the side wind environment. It directly affects the vehicle's operational safety and driving stability. On the basis of the domestic research on driving safety under the wind outside, this paper collects the unsteady flow parameters on the real road to establish a database of crosswind aerodynamic test conditions with regional characteristics, obtains and analyzes the air, vehicle and other parameters to analyze the change trend of crosswind, and uses a single factor to analyze the specific influencing factors of crosswind environment on vehicle driving safety. The driving simulation experiment in crosswind environment is carried out by using the dynamics simulation software CarSim to evaluate the driving safety and stability in crosswind environment. The air flow characteristics of the change of side wind environment have a great influence on driving safety and driving stability, and through the analysis and simulation of road measurement data, it is very important to analyze the factors influencing the safety of vehicle operation and the stability of driving.

To measure and analyze the transportation situation within the counties of Shaanxi province over a certain period of time, and to promote the province's future transportation planning and development, this paper proposes a reasonable layout plan and compares model rationality. The study uses entropy weight method to measure the transportation advantages of 107 counties in Shaanxi province from three dimensions: road network density, trunk line influence and location advantage in 2013 and 2021 and applies spatial autocorrelation analysis method to analyze the spatial layout and development trend of county-level transportation and conducts OLS regression on transportation impact factors and other influencing factors. The paper also compares the regression fitting degree of the GWR model and the OLS model. The results show that spatially, the transportation advantages of Shaanxi province generally show a decreasing trend from the Weihe Plain to the surrounding areas, and mainly exhibit high-high clustering phenomenon. Temporally, transportation advantages show an overall upward trend, and the phenomenon of spatial imbalance gradually decreases. People's travel demands have changed to some extent, and the demand for rapid transportation has increased overall. The GWR model regression fitting degree of transportation advantages is 0.74, which is higher than the OLS regression model's fitting degree of 0.64. Based on the evolution of transportation advantages, it is predicted that this trend will continue for a period in the future. To improve the transportation advantages of Shaanxi province, increasing the layout of rapid transportation can effectively enhance the transportation advantages of Shaanxi province.

With a permitted left-turn phase, left-turning vehicles irregularly change speed and trajectory to compete for the right of way with straight vehicles approaching from the opposite direction, which brings serious safety hazards to intersection operation. The data collected from three intersections in the high-tech development zone of Xi'an city, and the turning behaviors of left-turning vehicles were classified into three patterns: "no-interference, yield, and rush" based on the interactions between the left-turning vehicles and the straight vehicles approaching from the opposite direction. This study took post encroachment time as the quantitative indicators of the conflict severity, which was classified into 4 levels and the disordered logistic regression model was established to identify the significant factors affecting the conflict severity. The results showed that: the conflict severity of the left-turning vehicles in rush pattern is higher than that in yield pattern. The odds ratio of that the conflict severities were medium, serious conflict respectively increased to 11.598 and 3.348 when the vehicles in rush pattern. The arrival speed of the left-turning vehicles and signal phase were significant factors increasing the severities of conflicts, and the average speed difference and average headway are protective factors. The results of this study can provide a reference for traffic management departments to optimize traffic control schemes.

To recognize the key road sections that significantly impact the operational efficiency of the entire road network, a key road section identification model is developed based on an evaluation of their importance and weakness, taking into account the reliability of road network mobility. According to the definition of the mobility reliability of urban road network, mobility reliability of road section is calculated through the relationship between the saturation and mobility reliability of the road section. By analyzing urban road network structure, path and OD point-to-point reliability can be calculated. Considering the distribution of road network traffic flow, a mobility reliability model of road network is established. Important road sections are identified through an assessment that considers the connectivity of road sections, road’s distance, and the classification of each section. Meanwhile, combined with the mobility reliability, sensitivity analysis is conducted to determine weak road sections. By combining the importance and weakness of the road section, the key sections are identified. A complex network containing expressways, main roads, sub-main roads, and branch roads is used to verify and analyze. The findings demonstrate that the developed model and methodology effectively and accurately distinguish the critical sections. This method can provide reference and basis for the recognition of key road sections

In this paper, we analyze the numerical characteristics of stopping spacing and starting interval of vehicles at signal intersections, explore the relationship between forward vision and stopping spacing and starting interval, and analyze the impact of vehicles on road traffic under different stopping spacing conditions by simulation software. Firstly, the numerical characteristics of stopping spacing and starting interval are analyzed based on the data collected in the field; then, the correlation effect of vehicle type on stopping spacing and starting interval and its distribution characteristics are verified; finally, the motion process of vehicles stopping and starting at different stopping spacing is simulated in vissim simulation software to analyze the influence of different stopping spacing on road traffic. The results of the statistical analysis of the data showed that the differences in vehicle forward vision due to different vehicle types had a significant effect on the stopping spacing, while it did not have a significant effect on the start interval time; the simulation results show that as the stopping spacing decreases, the traffic capacity and average speed increases, while the delay and travel time decrease.

The development of intelligent transportation technology and the dual carbon policy have brought new opportunities for energy-saving driving. Energy-saving driving involves two problems, namely, real-time recognition of driving road conditions and driving styles for energy-saving driving evaluation, energy-saving vehicle speed calculation and energysaving driving guidance for non-energy-saving driving. Therefore, this paper designs a real-time recognition model of driving road conditions and driving styles and a global optimization energy-saving vehicle speed calculation model. Firstly, the urban roads in Tianjin and Hefei are selected to carry out the experimental analysis of the influencing factors of energy-saving driving. Secondly, through the selection of characteristic parameters, K-means clustering based on genetic algorithm optimization, and neural network training model based on genetic algorithm, the real-time recognition of driving road conditions and driving styles is realized in road test. Finally, based on the dynamic programming algorithm, a global optimization energy-saving driving strategy model is established to achieve energy-saving driving speed guidance for intense driving. The results show that using energy-saving speed guidance, the maximum reduction of vehicle driving energy consumption is 16.90 %, and the average reduction is 8.34 %.

To effectively improve the rescue efficiency after expressway accidents occur and ensure the smooth and orderly operation of traffic flow, this study analyzed the modes of vehicle following in a connected environment and defined the area adjacent accident point (accident site area and transition area) more accurately. The definition model of the area adjacent accident point was constructed according to the queuing length derivation process. Finally, by instance, the accuracy of the definition model of the area adjacent accident points was verified. Compared to the traditional model, the definition model of the area adjacent accident points can effectively reduce delays.

Along with the rapid development of electric vehicles, the continuous construction and continuous popularization of electric vehicle power supply facilities has become an urgent need. Nevertheless, unlike traditional fuel vehicles, with the increase in the ownership of electric vehicles, the charging behavior of large-scale electric vehicles connected to the grid in the future will have a non-negligible impact on the grid. Disorderly charging of EVs will increase the peak load of electricity consumption across the grid and exacerbate the peak-to-valley difference in load. In particular, the popularity of fast charging will increase the complexity and uncertainty of EV loads while increasing the load on the grid. In order to suppress or eliminate the negative impacts of EV charging, distributed PV plants, EVs, energy storage devices and their control devices can be combined and operated together. In this study, a distributed photovoltaic charging station with an installed capacity of 94 kW is built based on the 2020 weather data of Lanzhou area. Throughout 2020, the distributed PV charging station for electric vehicles can provide 103,749.19 kWh of clean electricity for electric vehicles, equivalent to 41.5 tons of standard coal. Not only can it ease the pressure on the power grid, but it can also continue to provide real clean power for electric vehicles.

The safe running of high-speed trains depends on the adhesion between the wheel and the rail. This paper suggests an optimal adhesion control for high-speed trains based on the use of the CKF (Cubature Kalman Filter) in order to completely exploit the adhesion between the wheel and rail and attain the ideal adhesion condition. The method allows for the approximative estimate of train speed using CKF and is based on the wheel-rail dynamics and high-speed train adhesion model. The extremum search algorithm without steady-state oscillation is used to build an optimization adhesion control, which eventually directs the train to run as closely as feasible to the peak adhesion point, overcoming the effects of external interference and estimating errors on the system. Finally, MATLAB/Simulink is used to create a simulation model of the adhesion control for the traction transmission of the CRH2 high-speed train. With an average adhesion utilization rate over 95%, the simulation results demonstrate the viability of the suggested approach, ensuring the effective and secure running of high-speed trains.

In order to study the characteristics of two-way narrow channel traffic flow under risk factors, the characteristics of twoway channel traffic flow in the risk area are analyzed, and the rules of two-way ship yielding and no yielding are proposed when there is a risk area in the channel. traffic flow in the risk area are analysed, and the rules of two-way ship yielding and no yielding are proposed when there is a risk area in the channel. The NaSch cellular automata is introduced according to the behaviour of the ship in the course of driving, and the rules of two-way ship yielding and no yielding are proposed. NaSch cellular automata is introduced according to the behaviour of the ship in the course of driving, and the cellular automata model of ship traffic flow simulation is established. The NaSch cellular automata is introduced according to the behaviour of the ship in the course of driving, and the cellular automata model of ship traffic flow simulation is established. Through the proposed navigation rules for ships in the risk area, a two-way narrow channel simulation rule based on cellular Through the proposed navigation rules for ships in the risk area, a two-way narrow channel simulation rule based on cellular automata considering the channel risk is proposed. simulation considering the channel risk can well simulate the two behaviours of ship yielding and no yielding in the channel with risk area.

With the rapid development of urbanization, as an important component of urban public transportation, the monitoring and statistics of passenger flow in urban comprehensive transportation hubs have become increasingly important. The design of passenger flow monitoring system of integrated transport hub based on big data is to meet the demand of passenger flow monitoring in such a complex environment. A passenger flow monitoring system of integrated transport hub based on big data is designed and proposed. The System integration integrates all kinds of passenger flow data sources, collects the number and distribution information of passenger flow inside the hub by studying the indoor positioning technology based on 5G mini base station, and realizes the monitoring of real-time passenger flow. The system design considers the real-time nature of data collection, accuracy of data processing, and stability of system operation, providing a comprehensive, accurate, and timely passenger flow monitoring tool for urban traffic management departments

In order to improve the transport efficiency of intermodal transport the suitability of intermodal transport in express cargo, this research takes intermodal express freight truck as the research object and newly develops a multi-objective optimization model. From the perspective of timeliness, cost and matching degree, the model finds out the railway vehicle model with the highest matching degree, which minimizes the total transportation timeliness and transportation cost of intermodal transportation and maximizes the relative matching degree of truck and cargo. In this research, the corresponding algorithm is designed to find the optimal solution of the model and find the optimal vehicle application scheme in different scenarios.

This paper presents an object detection method for traffic management based on the YOLOV7 model, using the bdd100k dataset for experimentation. The results show that the proposed method has good detection performance in traffic scenes. The main contribution of this paper is to improve the accuracy and efficiency of object detection in traffic scenes by applying the YOLOV7 model to the field of traffic management. The research results of this paper are of great significance for the optimization and improvement of traffic management systems. Future research can explore YOLOV7's performance on other target categories and consider algorithm optimizations to improve accuracy on new datasets.

To solve the traffic safety problems that arise during "mixed traffic" between autonomous and manual vehicles, this paper presents a harzard scenario simulation and safety assessment method based on autonomous driving crash reports. 197 crash reports of autonomous vehicles are selected from California Department of Motor Vehicles and used for autonomous driving hazard scenario extraction. Combining 37 pre-collision scenarios published by the National Highway Traffic Safety Administration, typical traffic safety scenarios are identified. A random forest model is used to identify key elements of typical autonomous vehicle accident scenarios. The CARLA microscopic traffic simulation platform is used to simulate two typical "mixed traffic" safety scenarios: Lead Vehicle Stopped and Following Vehicle Making a Maneuver. Three indicators - relative velocity, relative distance, and time-to-collision - are selected for constructing a safety evaluation suitable for autonomous vehicles. The research results show that in the Lead Vehicle Stopped scenario, when the following vehicle is 15 meters away from the front vehicle, the risk of collision is extremely high when the initial speed of the following vehicle is greater than 13.1 m/s. In the Following Vehicle Making a Maneuver scenario, when the initial speed of the following vehicle is greater than 12.5 m/s, the risk of collision with the front vehicle is also extremely high. The results can provide support for the safety evaluation of autonomous driving and the improvement of autonomous driving system.

In order to better understand the development statuses and future evolution trends of intelligent connected transportation, this paper firstly elaborates the impact of it on road traffic system, and then summarizes the development statuses at home and abroad from three aspects: academic research, industrial development and engineering applications, in which academic research is mainly discussed from vehicle environment perception, vehicle autonomous decision-making and vehicle control execution. Finally the future research focus and development trends are proposed, so as to provide reference for its subsequent development and applications.

The phenomenon of wheel idling, resulting from poor adhesion conditions, not only reduces traction utilization but also poses a safety risk to train operation. A unique peak adhesion point is observed in the wheel-rail adhesion characteristic curve. To maintain stable train operation at this peak adhesion point, it is crucial to enhance adhesion utilization. This paper proposes a nonlinear model predictive control algorithm for optimizing train adhesion control. Initially, the dynamic forgetting factor recursive least squares algorithm is employed to online identify track parameters. These identified online track parameters are subsequently utilized by the nonlinear model predictive controller to achieve optimal adhesion control. The aim is to position the train's operation point as closely as possible to the adhesion curve's peak, thereby achieving optimal adhesion control. A MATLAB/Simulink simulation model for vehicle wheel-rail adhesion is developed, and the simulation results demonstrate the superior effectiveness of the proposed adhesion control method when compared to linear time-varying model prediction adhesion control. Additionally, a higher average adhesion utilization rate is achieved.

The inadequate layout of intercity fast charging facilities leads to difficulties in intercity travel for electric vehicle(EV) users. We incorporate EV users' path selection behavioral preference factors into planning intercity fast charging facilities and propose a bi-level programming model. Firstly, the lower layer constructs a stochastic user equilibrium model for dynamic traffic in Logit form based on general regret theory, and the upper layer constructs a multi-objective planning model that considers minimizing the total annual cost of EV charging station operators and minimizing the average queuing time of EV users; secondly, the NSGA-II nested MSA algorithm is used for solving the bi-level programming model; finally, the feasibility of the model and algorithm is verified by the Nguyen-Dupuis transportation network example. The results show that our proposed model can meet the charging needs of different users, and can effectively reduce the queuing waiting time of EV users and alleviate their travel anxiety.

This paper builds a car following model in an intelligent transportation environment in order to thoroughly consider the driver's own characteristics and the impact of the environment on the stability of the traffic flow, taking into account the driver's memory effect and multi-vehicle information. According to the model's stability analysis and numerical simulation findings, the stability zone will grow as the driver memory effect lessens and more vehicle data is taken into account, while the amplitude of traffic flow density will steadily decrease. This implies that traffic congestion can be reduced while the consistency of traffic flow would be enhanced.

In order to provide a reference for EV travelers in charging route planning and charging station selection, a charging route selection model considering the impact of congestion on users' perception is established in this paper. Firstly, using the data obtained from the questionnaire survey, the MNL model is used to obtain the impact of congestion on users' perception of travel time. The results show that when the travel time is 10min, the user's mental travel time is 2.6 minutes longer than the actual travel time when the congestion degree increases by 1 level. Secondly, considering charging fee, parking fee, queuing time at the charging station and mental travel time from the starting point to the charging station and then to the destination, a charging route selection model aiming at minimizing the total charging cost is established. Taking the actual road network in typical areas of Beijing as an example, the charging route selection is carried out based on ant colony algorithm. The results show that the model is feasible and effective.

To study the characteristics of the mixed traffic flow composed of manually driven vehicles and connected automated vehicles in the intelligent connected environment, mainly exploring the relationship between flow, density, and speed, this paper constructs a basic graphical model of the mixed traffic flow in the intelligent connected environment. First, four different car-following models in the mixed traffic flow are analyzed, as well as the corresponding percentage of following modes under different penetration rates of connected automated vehicles. Based on this, the different following models, and the corresponding model parameters, are applied to derive the fundamental diagram model of mixed traffic flow in the intelligent connected environment. Finally, the simulation is verified by designing SUMO simulation experiments. The simulation results show that the flow-density data scatters obtained under different connected automated vehicle occupancy ratios are in high agreement with the theoretical curves, which verifies the feasibility of the model.

The emergence of mixed traffic flow involving Intelligent Connected Vehicles (ICVs) has increased the vital of studies on the impact of ICVs on road safety, especially freeways. It is also essential to comprehend the mixed traffic flow safety corresponding to the different penetration rates of ICVs in the section of the exit ramp of freeways. This paper presented an analysis of mixed traffic flow safety at the freeway exit ramp based on the vehicle behavior models calibrated by the genetic algorithm and simulation experiments by SUMO. The real dataset of the exit ramp was recorded in Hechizhai Interchange by an unmanned aerial vehicle. The simulation experiments were conducted with the different penetration rates of ICVs and the proportion of vehicles exiting the ramp. The results demonstrated that the higher penetration rate of ICVs would improve the safety of mixed traffic flow in the exit ramp of freeways, although the higher proportion of vehicles exiting the ramp would increase traffic conflicts and risks

In 2022, the National Bureau of Statistics reported a distressing toll of 61,703 lives lost and 250,723 injuries due to traffic accidents nationwide. This equates to a tragic average of one life lost every seven minutes, propelling our country to the top of global traffic accident statistics. Beyond the well-known factors contributing to accidents like fatigue and reckless behavior, there has been a concerning surge in accidents linked to underlying health conditions, notably heart disease and high blood pressure from poor health. In response, we've developed an innovative in-vehicle human health monitoring system. This product seamlessly integrates a pulse diagnosis module into the steering wheel, allowing real-time monitoring of pulse waveforms and critical health data, a pivotal component of our design. Our system also utilizes a trained opensource model to develop modules for monitoring and providing reminders for various driver behaviors, including closed-eye detection and phone call management. By meticulously analyzing the connections between the driver's heart condition, pulse diagnostics, and vehicle speed, our product generates personalized driving speed and timing recommendations that prioritize the driver's well-being. Furthermore, in emergencies, our system promptly notifies the driver's family of any detected physical abnormalities or the driver's whereabouts, instilling peace of mind during journeys. In essence, our product proactively mitigates the risks associated with traffic accidents stemming from driver fatigue, reckless driving practices, or irregular pulse patterns. It promotes safer driving habits and contributes to the early detection of certain health conditions, fostering safer roads and improved health outcomes.

In order to analyze the impact of ramp convergence in merging area on the dynamic characteristics of expressway mainline traffic flow under the mixed traffic flow environment composed of connected and automated vehicle (CAV) and humandriven vehicle (HV), this paper selects the intelligent driver model (IDM), the adaptive cruise control (ACC) car-following model, and the cooperative adaptive cruise control (CACC) car-following model to model the HV, the DCAV (degraded connected and automated vehicle), and the CAV following behavior, respectively. The proportions of different types of vehicles are determined by platooning intensity, and the effects of changes in penetration rate on the dynamic characteristics of mainline traffic flow on expressways are investigated in terms of operational efficiency, traffic safety, emission, and fuel. The results show that the increase of penetration rate is more significant to the improvement of free flow operation efficiency compared with congested flow, and can effectively inhibit the diffusion of blockage to the upstream. With the increase of penetration rate, the overall safety of mainline mixed traffic flow is improved based on time-to-collision (TTC) analysis. However, when the proportion of connected and automated vehicles is 0.7, the mainline mixed traffic flow locally has the highest risk of rear-end collision, and the time exposed time-to-collision (TET) and time integrated time-to-collision (TIT) achieve peak values. In addition, increased penetration rate significantly reduces exhaust emissions and fuel consumption in mixed traffic flows.

In order to explore the influence of the travel dynamics parameters on the emission characteristics of the gasoline direct injection(GDI) light-duty gasoline vehicle when driving in the plateau environment, a GDI light-duty gasoline vehicle was selected as the test object, and real driving emission(RDE) tests were carried out under different driving intensity to analyze the difference in vehicle emission characteristics. The results show that the robotic process automation(RPA) value of stroke dynamics parameters and ν·a_pos[95] has a significant linear positive correlation with NOx and CO₂ emission factors, among which NOx emission has the most significant correlation with stroke dynamics parameters, while CO emission has no significant correlation with RPA. When RDE test is carried out in plateau area, NOx emission level can be judged by stroke dynamics parameters.

With the rise of the social logistics industry, the vehicle routing problem has always been a hot topic for scholars to study. This paper focuses on the recent research advances in the last 5 years of the vehicle routing problem, which contains the basic vehicle routing problem (VRP), vehicle routing problem with capacity constraints (CVRP), vehicle routing problem with time windows (VRPTW), split delivery vehicle routing problem (SDVRP), dynamic vehicle routing problem (DVRP), other types of VRPs, and so on. By subdividing the above according to VRP characteristics and describing its solution method. Finally, future research trends and development trends of VRP are summarized.

Traffic flow prediction plays an important role in intelligent transportation systems(ITS), but is challenged by the spatiotemporal complexity of traffic flow connections. In order to integrate the traffic temporal correlation, spatial correlation and semantic correlation in road network models, we propose a deep learning framework, Temporal Multi-Graph Convolutional Neural Network(T-MGCN) for traffic flow prediction. Firstly, we identify several semantic correlations and encode the non-Euclidean spatial correlations and heterogeneous semantic correlations between roads into multiple graphs. These correlations were modeled by a multi-graph convolutional neural network. Next, a recurrent neural network model is used to learn the dynamic characteristics of the traffic flow to obtain temporal correlations; Finally, a fully connected neural network is used to fuse spatio-temporal correlations and semantic correlations

Urban rail transit network plays a key role in urban transportation system as an important travel mode to alleviate traffic congestion. This paper is based on complex network theory, considering the factors such as departure interval to travel time as the weight of network connecting edges, using Space-L method to establish a weighted complex network topology model of urban rail transit considering interchange stations, proposing comprehensive evaluation indexes to identify the key points of nodes in the network, and analyzing the topological characteristics and robustness of the weighted complex network under random and intentional attacks by taking Xi'an city rail transit network as an example. The topological characteristics and robustness of the weighted complex network under random and deliberate attacks are analyzed. The following conclusions can be drawn from the example analysis of Xi'an urban rail transit network: (1) when attacking the network nodes, the deliberate attack causes more impact than the random attack; (2) the comprehensive evaluation indexes cover the node degree and meson number, which can be effective in the identification of key nodes of the network; (3) the Xi'an urban rail transit network is still in the stage of construction When the key nodes are attacked, the network is easier to be paralyzed, and the network robustness can be further improved by adopting protection measures for the key nodes when constructing the urban rail transit network.

Unmanned Surface Vehicles(USV) path planning has important application value in the field of autonomous navigation. In order to plan a path with simple calculation, less time consumption and low complexity, an improved Probabilistic Roadmap(PRM) algorithm based on USV path planning is proposed. The randomly generated sampling points in the traditional PRM algorithm results in high computational complexity and insufficient flexibility and smoothness in the planned path. In order to optimize the generation strategy of sampling points and improve the smoothness of the path, the boundary box analysis technology and Catmull-Rom Spline interpolation of the path are introduced into the PRM algorithm to obtain I-NPRM algorithm. Using three different path planning schemes for simple, complex and special water environments and comparing them with the planning effectiveness of I-NPRM algorithm, NPRM algorithm, PRM algorithm and A* algorithm under three indicators of runtime, path length and number of inflection points. The experimental results indicate that the I-NPRM algorithm can optimize the sampling point generation strategy while reducing runtime and path length, as well as reducing the number of inflection points. The planned path has high smoothness, which is more in line with the motion constraints and maneuverability of the USV and improves the safety and stability of the autonomous navigation system

Lane detection plays a significant role in computer vision and autonomous driving applications. However, it encounters challenges under low-light conditions during nighttime, where images are darker and real-time processing is necessary. In this study, a nighttime lane detection method based on OpenCV is proposed. Initially, image contrast stretching is applied to enhance brightness as a preprocessing step. By adjusting the grayscale range of the image, lane line features can be captured more effectively. Median filtering and bilateral filtering are then employed to reduce noise interference in the image. To address potential left-right deviation caused by the camera, a novel dynamic Region of Interest (ROI) method is introduced. This method adaptively adjusts the ROI based on real-time image analysis, thereby reducing false detections and improving overall detection performance. Canny edge detection and Hough transform are utilized to locate the lane lines. Finally, lane fitting and drawing techniques are applied to determine the vehicle's position within the lane. Experimental results demonstrate the high accuracy and robustness of the proposed nighttime lane detection method, which incorporates enhanced images and the dynamic ROI approach. The method accurately identifies the vehicle's lane position, providing essential visual guidance for autonomous driving systems.

With the rapid growing of vehicle population, NOx has gradually become a key factor restricting the prevention and control of O₃ pollution in China. High-resolution vehicle emission inventories are essential for the development of scientific and efficient pollution prevention and control policies. In this study, a bottom-up approach is adopted to develop a high temporal and spatial resolution vehicle NOx emission inventory for Guangzhou in 2021 based on real-time road network hourly-scale traffic data detected by 5,198 AVI devices. The study further examines the prominent emission vehicle types, emission periods, and emission hotspots in different regions (11 administrative districts, central urban areas, and noncentral urban areas) to facilitate a detailed analysis and traceability of high-pollution events. The results of the study showed that the average daily NOx emission from vehicles was 196.15 t/day and the average daily emission intensity was 15.82 kg/(dayûkm). Diesel trucks, medium & heavy diesel trucks, and diesel trucks with China III or lower emission standards are the prominent vehicles with low occupancy and high emissions. In terms of spatial distribution, HDTs are the main contributors in the non-central urban areas, while heavy passenger vehicles and light duty trucks are the main contributors in the central urban areas, and buses also contribute a slightly higher percentage compared to the non-central urban areas. To meet the demand of freight transportation, the distribution of NOx emission hotspots shows a pattern of dispersion outward along the beltway network centred on Baiyun District. A more detailed traceability analysis of the high NOx emission contribution is used to assist in the development of pollution prevention policies.

In path planning for unmanned surface vehicles, the classic A* algorithm has the drawbacks of having more redundant nodes, more search nodes, and being close to obstacles. Combined with the navigation characteristics of unmanned surface vehicles, the A* algorithm is improved from four aspects. Firstly, to ensure the security of the planed path, the evaluation function considers the influence of obstacles repulsion fields. Secondly, when searching nodes, the expansion directions of the node are optimized while taking into account the target node's direction information. Thirdly, the shortest distance from the obstacles grids to the connection between the target node and the current node is computed. If the distance is less than the safety distance, the A* algorithm's search process is jumped out to increase search efficiency. Finally, to lessen the number of turning points, the redundant nodes of the planned path are finally deleted. The simulation results show that there are obvious advantages in the number of turning points, the security of the path, and the number of traversal nodes.

With the significant improvement of the national economic level, China's automobile ownership continues to rise. A complete station(pile) is an important guarantee for the popularization of vehicles. Reasonable planning of charging piles is crucial for promoting vehicles. As a means of transportation, vehicles have both road network load attributes and distribution network load attributes. Vehicles fulfill users' travel needs in the road network and obtain energy supply in the distribution network. They are a key link between the transportation and distribution systems, and an important component of the energy internet. This article first proposes a method based on dynamic traffic simulation to obtain the arrival rate of charging piles, and then calculates the time of waiting of vehicles. Then, taking the time of waiting for electric vehicle charging and the minimum cost of charging pile construction and operation as the goals of two stakeholders, a multi-objective planning model for highway electric vehicle charging piles is established. Once again, the multi-objective particle swarm optimization algorithm is used to solve the model and obtain Pareto frontiers. VIKOR is used to sort the frontiers and select the optimal solution. Finally, the rationality of the method is verified through numerical examples.

In recent years, with the rapid development of national economy and transportation, the number of motor vehicles has increased dramatically, and the resulting traffic problems have aroused widespread concern. However, traffic flow is easily affected by weather, historical traffic flow and human factors, so traffic flow has the characteristics of randomness and nonlinearity, so using a single forecasting method to predict traffic flow can not meet the needs of modern intelligent transportation system. In order to improve the accuracy of traffic flow prediction, a traffic flow prediction model TF-VMD based on Variational Mode Decomposition (VMD) and Transformer is proposed in this paper. We use VMD method to decompose the traffic flow time series into k modes, and use Transformer model to predict the traffic flow in the next period of time. Finally, we superimpose the prediction results of each mode to obtain the final prediction results.Through experiments on real traffic flow data, compared with other prediction methods, the prediction results are improved in average absolute error, average absolute percentage error and equality coefficient, which effectively verifies the superiority of the model.

Graph-based traffic flow predictioning is widely applied in traffic systems, where constructing intricate spatiotemporal correlation models from relevant time series data is imperative for comprehending the dynamics of the traffic system. The extraction of features from graphical data, coupled with the integration of time series data, serves to enhance the accuracy of traffic flow predictions. Additionally, the challenge arises when real traffic flow data often contains missing values. Predicting traffic in scenarios with missing data proves to be challenging, as existing traffic flow predictioning methods frequently lack the capacity to model dynamic spatiotemporal correlations in the presence of such gaps, leading to unsatisfactory prediction outcomes. This paper introduces the Adaptive Spatiotemporal Graph WaveNet-based Graph Convolutional Network (AST-GW-GCN) to tackle traffic flow predictioning. AST-GW-GCN comprises three independent components, each modeling short-term, daily, and weekly dependencies of traffic flow. Within each component, Gated Temporal Convolutional Network (TCN) and Graph Convolutional Network (GCN) serve as encoders, conducting spatial convolution to extract spatial correlations and temporal convolution to capture temporal correlations, thereby generating hidden features. The Gated Recurrent Unit (GRU) is employed to decode these hidden features and weigh the outputs of the three components, producing the final prediction result. The spatial convolution module establishes an adaptive adjacency matrix to overcome the physical constraints of the graph structure, facilitating improved extraction of hidden spatial dependencies within the data. Furthermore, experiments are conducted under various data missing patterns and missing ratios. The experimental findings, based on the PeMS08 real dataset, demonstrate that the proposed AST-GWGCN comprehensively captures spatiotemporal correlations in the data, outperforming baseline models in terms of performance.

A multi-objective planning on vehicle lane changing is constructed, the NSGA-II algorithm is used to find the Pareto feasible solution, and the High-D dataset is introduced for validation analysis. The results show that the driver's goal in lane changing is to pursue higher speed and efficiency, when the environment of vehicle lane changing affects the initial headway spacing, headway time distance, maximum acceleration, and maximum speed when changing lanes should 6.54 ~ 8.34m ,1.08 ~ 2.32s , 2 2.24 ~ 4.56 m / s² ,31.54 ~ 46.74 / s , when the vehicle changing lane can obtain higher speed and efficiency gains with lower vehicle carbon emissions.

Unmanned Surface Vehicle (USV) technology has been widely applied in various fields, due to the nonlinear, multiconstraint, high-dimensional nature of motion planning problems, as well as the dynamic characteristics of USV movement and the uncertainty of the environment, solving USV motion planning problems is highly complex and has an unpredictable failure rate. This paper focuses on the motion planning problem of surface USVs and studies the solution methods based on the optimal control model in modern control theory. By introducing the trapezoidal collocation method from the direct method, the optimal control formula describing the entire USV motion planning problem is fully discretized into a largescale non-linear programming problem (NLP), which is then quickly solved using an interior-point solver. In random scenario simulation experiments with different ship interaction conditions, the motion planning algorithm's solution time is within 250ms, basically meeting the practical engineering requirements. The highlights of this paper are as follows: 1. The optimal control method is used to ensure that the output control input is suitable for the USV power chassis; 2. The direct method is used to fully discretize the problem, transforming the optimal control problem, which essentially seeks the functional extremum in infinite space, into a NLP that can be quickly solved using mature mathematical tools to meet the low-latency requirements of USV working conditions; 3. The adoption of trapezoidal collocation discretization ensures that the constraints describing the control system in NLP are all first-order algebraic terms and their linear combinations, guaranteeing the low difficulty of solving NLP and effectively reducing the failure rate to within 10%.

This paper proposes a new traffic flow prediction method, which combines the grey relational analysis method with the long and short term memory network, and uses Python language to build the model. Taking California expressway network PEMS data set as the research object, the speed and occupancy of target detection points, as well as the traffic flow of surrounding detection points as influencing factors, are input into the traffic flow prediction model, so as to take into account the influence of multidimensional spatiotemporal factors on the prediction model. Specifically, grey correlation analysis GRA is used as the correlation analysis method of each detection point.^[1]-[4] By eliminating the traffic data with little or no correlation to the traffic flow of the target detection point, the data dimension of the input of the prediction model is reduced, and the efficiency of model training and prediction is improved. ^[5]-[7] Then, constructs a traffic flow prediction model based on stacked LSTM using several LSTM networks to capture the spatiotemporal characteristics of road traffic conditions that are strongly coupled in multidimensional traffic data types and massive data, thereby improving the effectiveness and accuracy of traffic state prediction.

The traditional logistics industry has a history of many years of development. However, with the changes of the times and the progress of technology, the county express logistics has gradually developed into a ' bridge ' connecting agricultural products into the city and industrial products into the village, which has become a key factor in promoting modern economic development and promoting rural revitalization. Aiming at the problems of high cost, poor service and difficulty in entering the home of rural express delivery, this paper puts forward the terminal distribution mode of joint distribution + crowdsourcing, and combines the actual geographical coordinates, uses genetic algorithm and MATLAB to verify the example, and obtains the most suitable location scheme.

In order to meet the demand of consumers to fresh goods and reduce the transport cost of fresh cold chain logistics, this paper firstly expounds the development status of fresh cold chain in our country. On this basis, the factors affecting the cost of fresh cold chain logistics in multimodal transport are analyzed, which are mainly the different costs caused by the different transport distance, transport mode, transport time and transport node selection. In order to reduce the fixed cost, transport cost of distribution vehicles, node transport cost, freight damage cost in the process of refrigerated fresh food transport and energy cost consumed by refrigerated goods in the process of distribution brought by various influencing factors, the cost optimization model of fresh cold chain logistics multimodal transport was established and solved by genetic annealing algorithm for example. The multimodal transportation path with the optimal cost within the time specified by the customer is obtained, which effectively reduces the penalty value in the distribution process and reduces the transportation cost.

In view of the current traffic congestion problem, if only rely on the traditional traffic light indication is no longer able to deal with road traffic congestion, so we need to adjust the traffic light indication time in real time according to the realtime traffic conditions on the road, so as to achieve the effective unblocking of intelligent traffic lights. This paper presents the design of intelligent traffic light system based on genetic algorithm to optimize Elman neural network. First of all, this design will take STM32F103 single chip microcomputer as the core to build the hardware circuit, including the collection of traffic flow, automatic control and feedback of the closed-loop control system. In terms of software design, this design builds a system program and processes the collected traffic flow information by optimizing Elman neural network algorithm based on genetic algorithm. According to the obtained results, the display time of traffic lights in the next cycle is changed, so as to achieve the purpose of regulating traffic flow and improving traffic.

To alleviate the conflict problem between right-turning vehicles and crossing pedestrians at signal intersections, this paper uses the right-turn delayed release behavior control strategy to analyze the intersection's pedestrian-vehicle conflict characteristics, selects the delay and the number of conflicts as evaluation indexes for efficiency and safety, and uses the entropy weighting method for weighting to establish comprehensive indexes, so as to explore the changes of comprehensive index scores under different pedestrian and right-turning traffic, and finally in VISSIM simulation software, through the COM interface using python programming language for its secondary development, build the intersection conflict environment for simulation experiments, and analyze the simulation results. The results show that when the rightturn flow is between 200 pcu/h and 500 pcu/h and the two-way pedestrian crosswalk flow is not less than 400 ped/h, the composite index scores are improved after setting the right-turn delayed release phase.

This study addresses the location problem of electric vehicle charging facilities by constructing a bi-level programming model and designing a hybrid algorithm for its solution. The upper planning model aims to minimize the construction and operation as well as maintenance costs of charging facilities, while the lower planning model focuses on minimizing the charging costs for users. Traditional heuristic algorithms often encounter issues of getting trapped in local optima and yielding low solution accuracy when solving location models. Therefore, this study proposes a hybrid algorithm, namely the particle swarm-grey wolf hybrid algorithm(IPSOGWO), based on the particle swarm optimization algorithm (PSO) and an improved grey wolf optimization(GWO). This algorithm incorporates a nonlinear convergence factor to enhance the search and development capabilities of GWO and combines the two algorithms by improving the position update formula of PSO. Experimental results from optimizing four benchmark functions demonstrate that IPSOGWO outperforms traditional heuristic algorithms in terms of convergence speed and accuracy. Based on this hybrid algorithm, the location model is effectively solved. The results reveal that the proposed algorithm and model successfully reduce the overall location cost and provide novel approaches and insights for the scientific placement of electric vehicle charging facilities.

The wave of Industry 4.0 drives the development of intelligent manufacturing. A modern precast girder yard is a complex system that collects multiple elements such as intelligent sensing equipment, construction personnel, facility instruments, and data processing platforms, etc. The existing management platform has problems such as scattered management processes, poor integration, and poor delivery efficiency, etc. How to promote the interaction and collaboration of complex systems and increase the yield of precast girder yards is the current challenge for enterprises and researchers. This paper adopts a model-based system engineering approach to model the system requirements, system environment, system architecture, subsystem interactions and functions of the smart girder yard, and combines digital twin technology to build a digital twin of the physical girder yard, thus realizing visualization of girder yard production, construction simulation and intelligent scheduling decisions. Through system verification, the application of the smart girder yard system reduces the interference of human factors in the production schedule of the girder yard, improves the production efficiency, and meets the requirements of digital control of all elements of the production schedule.

Electric logistics vehicle route planning plays a crucial role in the operations of urban distribution service providers. By optimizing vehicle routes, not only can significant reductions in operating costs be achieved, but also enhanced customer satisfaction. However, the vehicle path problem is a classic NP-hard problem in the field of combinatorial optimization. Therefore, finding an efficient and accurate solution holds great practical significance. In response to the real-time operational requirements, this paper proposes a novel approach that improves the graph convolutional neural network model with reinforcement learning. This approach aims to learn effective strategies for generating vehicle route schedules. Through comprehensive experiments conducted on real geographic data, several key findings have been obtained: 1) Explicitly considering edge features within the model is essential; 2) Employing a joint learning strategy can expedite training convergence and improve solution quality, and 3) Comparative evaluations against existing algorithms in the literature demonstrate the superior performance of our proposed model.

The Rear Cross Traffic Alert (RCTA) system can effectively keep drivers away from the danger of rearward collision, but the level of the RCTA system varies from vehicle to vehicle. There is a lack of a common test method to verify whether the RCTA system is effective. In this paper, by analyzing the functional principles and solutions of RCTA, the testing and evaluation method of the RCTA system including vehicle, pedestrian Adult, pedestrian child, and bicycle scenarios is designed and built. In addition, three different vehicles were selected for validation tests, and the results showed that the designed scenarios can well validate the RCTA system, which is meant for testing and evaluating the RCTA system of vehicles.

DMI can reflect the operation status of on-board ATP of the train control system in real time, and the recognition of its display content is an important prerequisite for realizing the automatic test of on-board ATP. The existing identification DMI technology has incomplete identification types and low detection speed. In order to achieve a better detection effect and a certain degree of wide applicability, a DMI information recognition method based on the YOLOv8 instance segmentation model is proposed. This method improves the segmentation and classification accuracy of DMI interface information by improving the training strategy of the instance segmentation task, completes the extraction of text information by combining OCR technology, and achieves high-precision segmentation of the most restrictive speed profile by combining the Douglas-Peucker algorithm. The experimental results show that the method proposed in this paper can accurately extract all the useful information of the DMI interface, meet the needs of laboratory testing, and lay the foundation for automatic testing.

To address the low-carbon multimodal transportation path planning problem, this paper firstly constructs a model of multimodal transportation path planning considering carbon trading, which minimizes total transportation cost as the objective function. A reinforcement learning-based genetic algorithm is designed to solve the model, and the Q-learning algorithm is implemented to adjust the crossover probability and mutation probability in the genetic algorithm to optimize the ability to avoid getting trapped in the local optima of the algorithm. Finally, arithmetic experiment verifies that the reinforcement learning-based genetic algorithm has a superior performance and the method can effectively reduce the transportation cost of multimodal transportation, improve transportation efficiency and reduce carbon emission.

Accurately understanding the timing characteristics of urban rail transit (URT) passenger flows is helpful in improving the operational efficiency. A deep learning-based approach is used in prediction in URT systems, leveraging the timing characteristics of stations to classify them. Firstly, the dynamic time warping (DTW) is employed to quantify the dissimilarities, while the K-means algorithm is utilized to categorize stations based on the timing attributes of passenger flows. Secondly, in order to mitigate the impact of data noise, the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) approach decomposes passenger flow data. Finally, an improved time series transformer model is proposed for deep learning prediction of passenger flows at different types of stations. The effectiveness of the approach is confirmed through the analysis of Xi'an Metro's passenger flow data, revealing the presence of four station classifications: occupation-residential balance type, business office type, leisure and entertainment type, and dense residential type. Compared with the Auto Regressive Integrated Moving Average (ARIMA) model, the Support Vector Regression (SVR) model, the Long Short-Term Memory (LSTM) model and Transformer, Mean Absolute Error (MAE) of prediction results in different types of stations is reduced by 8.94%~34.32%, Root Mean Square Error (RMSE) is reduced by 8.30%~36.50%, and Mean Absolute Percentage Error (MAPE) is reduced by 13.77%~38.84%. The accuracy of prediction for different stations using the proposed method is superior to the benchmark models.

Unmanned aerial vehicles (UAVs) have widely sprouted up as a potential application because of their high degree of mobility, adaptability and availability for deployment on-demand. Considering that the ground base stations (BSs) may be damaged in emergency situations such as natural disasters, resulting in the failure to serve ground users (GUs), this paper explores a UAV acting as an aerial BS to offer service to GUs. Maximizing the minimum throughput of GUs is selected as our aim in order to guarantee user fairness, and an optimization problem is subsequently formulated. For dealing with the non-convexity of the problem, we put forward a high-efficiency of algorithm by utilizing block coordinate descent (BCD) and successive convex optimization (SCA) techniques for an effective solution. In comparison to other benchmark schemes, simulation results demonstrate that our design yields a considerable gain in throughput.

Ship navigation in obstacle environments using traditional manual route plotting methods is not only time-consuming and labor-intensive, but also results in highly inaccurate route drawings. In case of changes in obstacle positions, the entire route needs to be redesigned and redrawn. Additionally, manually plotted route maps are not easily preserved and have limited applicability. In order to overcome the limitations of manual route plotting, a ship route planning method based on an improved ant colony algorithm is proposed. Simulation of the improved ant colony algorithm is conducted to obtain the optimal ship route in obstacle environments.

In existing non-real-time response customized bus planning research, only the direct line planning problem of passengers with a single starting and ending point, vehicle type and sub-path is generally considered, while ignoring the development of diversified line operation modes. Based on the theory of conventional bus planning, this paper refers to the vehicle routing problem (VRP) model, combines with the characteristics of customized bus operation, constructs a customized bus line planning model considering multiple starting and ending points, multiple vehicle types, passenger mixed loading and multiple path selection operation modes, and uses genetic algorithm to solve the model. Finally, an example is used for analysis. The results show that the average number of stops in this model is small and the load factor is above 70%, which proves that the model can obtain a more ideal line operation plan. Compared with the driving scheme of a single sub-path that only considers the shortest time and the shortest driving distance, considering multiple sub-paths can select an optimal solution that balances the driving distance and driving time. The research results of this paper can enrich the operation mode of customized buses, increase the flexibility of vehicle allocation and line integration of operation enterprises while expanding the coverage of passenger travel demand, and effectively improve the response volume of passenger demand and the degree of enterprise benefits.

Aiming at the problem that the trajectory direction cannot be distinguished from the historical data of ship Automatic Identification System (AIS) in the traditional clustering process, a DBSCAN clustering method based on the combination of improved Euclidean distance and cosine similarity is proposed. To measure the differences between arbitrary trajectory clusters in different directions, the improved DP algorithm is used to quickly and accurately extract the ship trajectory feature points, and maintain a high similarity with the original trajectory. The historical AIS data of the waterway near Port area of Zhoushan was selected to analyze and verify the algorithm. The results show that the proposed method can achieve the expected clustering effect, and the results are consistent with the actual situation, which has certain reference value in channel planning, navigation forecast, ship dynamic monitoring and other aspects.

Blind Area Detection System (BSD) plays a positive role in reducing the occurrence of car accidents. This paper selected the test data of 16 cars passing through the C-NCAP(2021 edition) BSD C2C overtaking scene, and the box diagram is used for statistical analysis of the data. It is found that the warning time TTC of vehicles equipped with a BSD system is reasonable at present, which plays a safety warning role and will not cause too much discomfort to users. However, the alarm timing of the BSD system of a few vehicles is not suitable enough, which still needs to be improved. In addition, after the target vehicle crosses the test rear, the BSD system should use a more urgent alarm method to remind it.

As a modern industry, smart logistics is of great significance for promoting the transformation and development of the green economy, optimising the government's macro-control, and creating an upgraded version of China's economy. In order to improve the efficiency of logistics, improve the low information transparency and poor security of urban logistics industry, according to the current situation of the logistics industry, based on the technical characteristics of the blockchain, the model of the smart logistics platform is constructed, and then the smart logistics platform is designed from the technical architecture of the smart logistics platform, the security layer, the business application layer, the data layer, and the data layer, and finally, this paper clarifies the aspects of reducing the cost of logistics operation, improving the transparency and traceability of data information, strengthening the information security, and optimising the supply chain management. Finally, this paper clarifies that blockchain has a significant impact on the development of smart logistics from the aspects of reducing logistics operation cost, improving data information transparency and traceability, strengthening information security, and optimising supply chain management.

A vehicle routing optimization model was created with the aim of reducing delivery costs by taking into account elements including simultaneous pickup and delivery, carbon emissions, and time windows. The vehicle routing optimization solutions for scenarios with and without consideration of carbon emissions, with and without simultaneous pickup and delivery were separately calculated using a genetic algorithm on a conventional dataset. According to a comparative examination of the data, it may be successfully controlled and improved product circulation rates, logistics delivery costs, and carbon emissions by taking into account both simultaneous pickup and delivery and carbon emissions components. This study offers useful information for making decisions to improve logistics delivery to consumer locations.

In order to reduce the distribution cost of terminal logistics and improve customer satisfaction, a multi-objective vehicle and drone joint distribution route optimization model is proposed to minimize the total cost of distribution and maximize the average satisfaction of customer under the constraints of time window, vehicle load, drone load and the longest flight distance. The pareto solution set is obtained by the elitist non-dominated sorting genetic algorithm (NSGA-Ⅱ), then the entropy weight method is used to determine the weights of each objective function, and finally the optimal distribution scheme is obtained. The experimental results show that the joint distribution model of vehicle and drone can effectively reduce distribution cost and improve the customer satisfaction compared with the traditional vehicle distribution mode.

Airport passenger flow forecast is the basis of revenue management for airlines. Accurate forecast model can help airlines develop more accurate sales strategies to reduce costs and maximize revenue. Since BP neural network model has the problems of slow convergence speed and easy to fall into local optimality when predicting passenger flow, improved particle swarm optimization algorithm is used to optimize BP neural network model and establish PSO-BP neural network prediction model. The passenger flow data of Beijing Capital Airport is used to carry out the experiment. The results show that compared with BP neural network, PSO-BP neural network can effectively improve the accuracy and stability of the forecast, and provide a new way of thinking for airport passenger flow

Demand identification and route planning are important considerations for the development of customized buses. In the demand identification section, we first identify potential scenarios for customized buses based on the stability and high frequency of passenger flows using taxi order data, then select OD areas in "many-to-one" scenarios and determine passenger demands information between each OD area. In the route planning section, based on spatio-temporal clustering of passenger demands to determine candidate stations for customized buses, customized buses route planning model is constructed with the objective of minimizing passenger travel costs and maximizing operator profits, while considering constraints such as passenger travel time windows and vehicle capacity. In order to balance passenger satisfaction and operator profitability, a dynamic fare function is introduced to calculate the operator's ticket revenue instead of using a fixed fare. The results of the validation using taxi order data from Xi'an demonstrate the feasibility and effectiveness of the proposed customized bus route planning design method in "many-to-one" scenarios. It enables pre-assessment of route profitability, providing decision support for the establishment of customized bus services.

The research topic is quantifying the relationships between air quality and transport. The purpose of this study is to collect and analyse air quality as well as the social and economy data which indicates the development of transportation and urbanization. The mainly methodology behind this paper is using multiple linear regression to analyze the relationships among air quality, the development of transport in high speed railway and population associated with more than 300 Chinese cities’ geographical distribution. There are various kinds of tools used in this research. When processing the data, SQL database and Excel is used to manage all the database and make it the formation which can be easily accessed. The multiple linear regression in SPSS is used when analyzing the relationships in all kinds of data. At this stage, multiple zero-one dummy variables are defined to help with the linear regression analyse. It can be seen that air quality is strongly fluctuate with time change in a year, also has very high correlation with the geographical distribution. Generally speaking, air quality index is associated with various kinds of variables as time change, geographical distribution, social and economy development which indicates urbanization as well as different air pollutants. This paper is to analyze the relationships among all of them and observe how the air quality fluctuate.

Railroad transportation is an efficient and economical way of cargo transportation. Due to the imbalance of supply and demand in the railroad freight market, various factors have complex forms and different degrees of influence on the freight volume, which makes the forecasting of the railroad freight volume complexity and non-linear characteristics. This paper combines grey relation analysis and BP neural network to forecast the national railroad freight volume. Firstly, the data of the factors affecting the railroad freight volume from 2012 to 2022 are selected, and the grey relation analysis is used to obtain the relatively high correlation of the four influencing factors, namely the proportion of highway freight volume, the proportion of water freight volume, the mileage of railroad operation and the value added of the primary industry, which are used as the input layer of the BP neural network; then, according to the corresponding relationship between the influencing factors and the volume of railroad freight volume, the model is trained; finally, the model based on the model of railroad freight volume is trained; and the model based on the model of railroad freight volume is trained. Then, the model was trained according to the correspondence between each influential factor and railway freight volume; finally, the grey relation-based BP neural network model was compared with the traditional BP neural network model. Finally, the grey relation-based BP neural network model is compared with the traditional BP neural network model. The results show that the grey relation-based BP neural network can not only get better solutions, but also shorten the training time.

With the acceleration of urbanization, the rapid increase of rail transit passenger flow has brought great pressure on the operation and management of rail transit stations. In this paper, the Xizhimen subway station in Beijing is a research object. The survey and research focus on the large passenger flow at the A1 entrance, A1 exit, and Line 2 in the Xizhimen subway station during the morning peak hour. Based on the field data of passenger flow in various parts of the railway station, the spatial-temporal distribution characteristics of large passenger flow during the morning peak hour were obtained. In addition, based on these characteristics, the corresponding evacuation time models in each part of the railway station established in accordance with the three directions of passengers. The research results can provide the basis for the design of the subway station and the organization and management of rail transit.

This paper presents an optimization analysis of the cold chain logistics network structure based on genetic algorithms. Firstly, the theoretical overview of the cold chain logistics network is introduced, including the concept, characteristics, and operation process of cold chain logistics, as well as the functional effects of logistics networks. Then, the structure of the cold chain logistics network is studied, including its composition and link analysis, and the formation mechanism of the cold chain logistics network is discussed. Subsequently, the genetic algorithm and its basic ideas and solution steps are explained in detail, and a genetic algorithm-based optimization model for the layout of cold chain logistics network nodes is constructed. The feasibility and effectiveness of the model are verified through case analysis. Finally, the research results of the paper are summarized, emphasizing the importance of optimizing the structure of the cold chain logistics network to improve operational efficiency. The research findings of this paper are of great significance for promoting the development of the cold chain logistics industry.

This paper investigates the methods for intelligent logistics distribution decision-making and designs a smart logistics distribution decision system based on Spring Boot. In terms of warehousing and sorting, the sorting process is optimized by establishing a work environment model and computational rules. For freight scheduling, a comprehensive cost estimation model and a network location model are developed, and genetic algorithm and improved local search algorithm are used for freight scheduling. In the system design, the Spring Boot framework is used to implement functional modules, including warehousing and sorting module, freight scheduling module, and data analysis module. The system has been tested for compatibility, response time, and user interface, demonstrating good performance and user experience. The outcomes of this research provide references for improving logistics distribution efficiency and reducing costs, and have practical application value. Future research can focus on further optimizing algorithms, expanding system functionalities, applying the system to actual logistics fields, and promoting the development of intelligent logistics [1].

By taking advantage of the low cost and high flexibility of UAV, the "truck-UAV" joint distribution mode is considered to be introduced into the logistics "last kilometer" distribution. Considering the limitations of UAV such as endurance and load capacity, and considering the customer demand exceeding the load of UAV and time window requirements, the "truck-UAV" joint distribution optimization model is built with the minimum total operating cost as the goal. According to the characteristics of the problem, spatio-temporal clustering is introduced, and a two-stage solution method is proposed. The effectiveness of the algorithm is verified by different scale examples. The results show that the proposed distribution scheme can effectively reduce the logistics cost of the traditional distribution mode, and provide a theoretical basis for the optimization of the distribution scheme of logistics enterprises. The research results enrich the research of truck - UAV joint delivery problem with time window.

Transport structure adjustment is an important means for collaborative emission reduction of greenhouse gases and conventional pollutants. In this paper, for the regional freight structure optimization, a multimodal freight network is built, the transport utility function is analyzed, a model for multi-objective regional freight structure optimization is built, Beijing-Tianjin-Hebei region is taken as an example, and the factors (e.g., the freight demand in the future years, the policy orientation for freight development, the development scales of various freight modes in the future) are taken as the constraint conditions to determine the feasible region of freight structure, build a model for multi-objective freight structure optimization with the minimum carbon emission and the maximum transport utility, and obtain the optimal freight structure in the year of carbon peaking and carbon neutrality goals, thus providing feasible ideas and methods for the sustainable development of the transport system.

In order to improve the accuracy of short-term traffic flow prediction, a combined prediction model based on Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and Seagull Optimization Algorithm optimized Long-Short Term Memory neural network (SOA-LSTM) was proposed. Firstly, CEEMDAN is used to reduce the interference of data noise, decompose the original passenger flow sequence into multiple characteristic components and residual, and optimize the hyperparameters of the LSTM neural network using SOA. Each feature component is predicted, and the prediction results of each IMFs are summed to obtain the short-term entrance and exit passenger flow prediction results of each station. Extracting four consecutive weeks inbound passenger flow data of Beijing's urban rail transit network in 2019, and validated and compared the proposed short-term passenger flow prediction model. The experimental results demonstrate that compared with the transfer passenger flow prediction results of LSTM, ARIMARBF, LSTM-BP, and SOA-LSTM models, the proposed CEEMDAN-SOA-LSTM model has lower RMSE, MAE, and MAPE error values than other models, with R2 increased by 1.63%, 1.2%, 1.18%, and 0.72%, respectively.

The paper aims to optimize the functional area layout of Weihai cold chain logistics base. On the basis of the functional positioning of the cold chain logistics base, the paper optimizes the layout of the functional area and determines the relative position of each functional interval by the design of the specific functions of the cold chain logistics base. At the same time, the system layout planning method (SLP) is used to analyze the correlation degree of logistics relationship and nonlogistics relationship of each functional interval, and explore the possibility of optimal design of the base. By constructing the system layout model of Weihai international cold chain logistics base, using the SLP method to optimize the design, clarifying the layout points and principles of each functional area, and putting forward the corresponding layout scheme, so as to make the facility layout of the base more scientific, reasonable, efficient and economic. This paper analyzes the practical feasibility and effect of the optimized design scheme, and provides useful reference and support for the development of Weihai cold chain logistics base.

The reasonable layout of cold chain logistics base facilities in Jinan is an important path to promote the healthy development of cold chain logistics. The location selection of the cold chain logistics base must be closely related to the economic development of the surrounding area and the industrial planning of the city, and cannot be separated from the surrounding industrial economic environment. At the same time, the cold chain logistics base is close to the consumer market, which can reduce the transportation cost, expand the radiation radius of the base and increase the purchasing power of consumers. This paper first explains the cold chain logistics base in Jinan City, and then analyzes the site selection of the base based on Analytic Hierarchy Process and Fuzzy Comprehensive Evaluation Method.

Due to the impact of holidays and promotions, e-commerce users' order volumes fluctuate significantly. In addition, sudden events such as epidemics and earthquakes may lead to the temporary or permanent closure of logistics facilities, resulting in parcels being urgently diverted to other logistics locations. These factors will affect the number of parcels transported on different routes and the number of parcels handled by different logistics locations. To address these issues, this paper constructs an emergency transportation and structural optimization model for logistics, focusing on the relationship between logistics facilities and transportation routes in e-commerce logistics networks. By applying mathematical models and algorithms, the volume of goods in each logistics facility and route is predicted to propose an optimised plan for transportation and sorting. The proposed model is validated and implemented using real e-commerce logistics data and existing statistical data. The results provide important theoretical guidance and decision support for ecommerce logistics networks, helping to reduce operational risks and improve operational efficiency.

Wheelset as a key component of the running part of railway vehicles, the geometric relationship between wheel and rail is complex, and the collision vibration of wheel and rail has a great impact on the stability of vehicle operation, which will affect the driving safety in serious cases. Therefore, this paper mainly studies the dynamic behavior characteristics of wheel-rail system under snake excitation. In the case of ignoring the rolling and shaking motion of the wheel-set, the transverse and vertical motion of the wheel-rail system is considered to be weakly coupled, and the vibration model of elastic collision of the wheel-rail system of a quarter vehicle is established to study the nonlinear dynamic characteristics of the wheel-rail system of the vehicle.The analytical solution is obtained by using semi-analytic method, and then the fourth-order Runge-Kutta method is applied to numerical simulation by MATLAB software. The simulation parameters of vehicle wheel-rail system are fine-tuned, and abundant nonlinear dynamic characteristics are found.

Air braking is the main form of braking for freight trains. With the rapid development of heavy-duty freight trains, the running speed of trains continues to increase, and the number of train formations and axle loads are constantly increasing. In order to cope with more complex brake environments, the braking system needs to further study the error in controlling train air pressure and methods to improve control accuracy. This article establishes a model of the entire vehicle air braking system based on AMESIM. Based on the different braking stages of freight trains as the theoretical basis, the train pipe constant pressure, air temperature, brake system pipe lengths, and air distribution valve parts parameters can be adjusted to obtain a model that is more in line with the actual situation. In addition, this paper confirms the accuracy of this model through mutual verification with brake tests.

The TOA ranging model is a model that calculates the distance between two points using the time difference of signal propagation. Based on the TOA model, several signal receiving base stations are established. By measuring the speed and time of the UAV receiving signal, we predict the location data of the UAV most suitable for delivering logistics parcels, shorten the transportation time, reduce the transportation cost, so as to achieve the maximum benefit. By analyzing the initial position data of the UAV before delivery, it can be estimated that it delivers packages in a nonlinear distributed path. Using unconstrained optimization model, using the nonlinear least-squares legal, into the drone delivery before the initial position of the package data, obtained about the position of the unmanned aerial vehicle delivery of the package after the nonlinear function, so as to achieve the drone distribution of location monitoring.

To manage the expenses associated with overseas warehouses and enhance customer satisfaction with their delivery, this paper establishes a bi-objective modelwhich aims to minimize total expenditure and optimize customer contentment. The model includes expenses related to constructing warehouses, transportation, tariffs, and storage.In this paper, 10 European overseas warehouse alternative points are selected to deliver goods to 8 European demand points, taking Jingdong's self-built overseas warehouse as an example, assigning weights to the two objectives of total cost and customer satisfaction, and solving the optimal warehouse building point using LINGO software. Finally, it is found that by changing the weights of the two objectives, different site selection results can be obtained, and enterprises can determine the weights to suit their requirements, offering suggestions on constructing foreign warehouses.

With the process of globalization and the continuous development of the economy, the role of the logistics supply chain in corporate operations has become more and more prominent, and its optimization has become a core topic of modern management. This research aims to explore the optimization strategy of logistics supply chain through machine learning technology. First, based on an in-depth analysis of machine learning theory and supply chain processes, a specific machine learning model adapted to this scenario was selected, and empirical research was conducted to explore the problems and challenges that may be encountered in practical applications, and for this purpose, specific suggestions were put forward. sexual optimization plan. It is expected that through the application of machine learning, efficient, accurate and flexible management of the logistics supply chain can be achieved, bringing substantial benefit improvements to modern enterprises.

Traffic flow is characterized by nonlinearity, volatility and randomness. To further improve the accuracy of short-term traffic flow prediction, a combined short-term traffic flow prediction model (CEEMDAN-CNN-LSTM) based on complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), convolutional neural network (CNN), and long short-term memory network (LSTM) was used. The model utilizes CEEMDAN to decompose the original traffic flow data into k smooth intrinsic mode functions (IMFs), inputs each modal function into the CNN-LSTM model for prediction respectively, and aggregates and accumulates the predicted values to obtain the short-term traffic flow prediction results. In the model, CNN is used to better capture the spatial characteristics of the traffic flow. The experimental results show that the combined prediction model has a high prediction accuracy compared to the ARIMA, LSTM, CNN-LSTM, CEEMDAN-LSTM, and EMD-CNN-LSTM models with reductions of 50.7%, 44.6%, 39.7%, 20.7%, and 9.7% in terms of MAE, respectively.

The port logistics system is the core of port cargo circulation. In the context of the new development pattern of domestic and international dual circulation, it is worth considering how to study the coordination of various production and operation links at the system level, in order to achieve efficient operation and smooth logistics supply chain. To analyze the synergistic effects and effects of different measures on the port logistics system, the article uses system dynamics methods to construct a port logistics system collaborative model. Taking Shanghai Port as an example, the article simulates and analyzes the collaborative situation of the port logistics system under different measures and resource inputs. The results indicate that comprehensive measures are more effective than single measures for system collaboration; Among the three comprehensive measures of optimizing production organization, strengthening innovation and strategic cooperation, optimizing production organization has the best effect on alleviating port congestion, and innovation and strategic cooperation have significantly improved the comprehensive service level. Port enterprises can adjust the proportion of resource investment and allocate resources reasonably based on time, port congestion level, and available funds.

In this paper, from the perspective of improving direct passenger flow, reducing line running time and line standardization, we take the maximum direct passenger flow per unit time as the optimization objective, take line running time and nonlinear coefficients as the constraints to construct the objective planning model of the bus line network, and put forward a kind of "second path" concept for the solution algorithm of this paper. A "second path" concept is proposed for the solution algorithm of this paper, and finally, The ant colony algorithm is solved by MATLAB. Based on the principle of "laying line by line and optimizing the network", the line network is laid out. Taking Jinan Laiwu urban bus as an example, the current 21 routes are simplified to 20, which saves unnecessary capacity resources, but the efficiency of direct passenger flow is increased to 94.7%, which is 7.7% higher than that before the optimization, and the running time of the routes is shortened, and the non-linear coefficients and unbalanced coefficients of the passenger flow are significantly reduced, and the line specification is significantly improved, which is a big improvement than that before the optimization. The results show that the proposed optimization model and solution algorithm can get a better bus line network, and then improve the operation efficiency of urban buses.

Express terminal logistics distribution is a logistics activity in which a logistics service provider delivers express parcels directly from the last distribution outlet to consumers in order to meet the end customers of the logistics link. In recent years, with the sharp increase in the amount of express delivery, because the residents and students are in class, it is impossible to directly deliver the express delivery to the customers, and the community unit also limits the security management of express delivery, and gradually there have been contradictions. For random entry and exit, it is necessary to perform multiple express delivery before frequent delivery, which has a great impact on the efficiency of terminal logistics and customer satisfaction. In order to reduce the difficulty of terminal logistics, this paper proposes a terminal circulation method established between the urban circulation center and the customer's terminal circulation network, that is, from the urban circulation center to the terminal circulation network and the customer's circulation method. The hub station between the enterprise distribution center and consumers will cover the entire community, office buildings, schools and other distribution terminals. Customers can freely choose the method and time of receipt to reduce the costs incurred when the two parties deliver the package and conduct the transaction. Time is more flexible.This paper focuses on the development status of terminal logistics distribution in domestic express delivery industry, and gives an overview of terminal logistics distribution and immune algorithms, and analyzes the significance of terminal delivery in express delivery companies. At the same time, for the problem of the location layout of the terminal distribution network of the express delivery company, a location model considering the distribution volume and distance is established, and the problem-solving ideas and steps of the immune algorithm are introduced. Through the development status of SF express company as the background of site selection research, master the relevant data of SF express company's terminal distribution site selection, combined with the development trend of terminal distribution, and apply immune algorithm ideas to solve the results in Matlab environment. Finally, it puts forward reasonable suggestions for the location selection of the terminal distribution network of SF Express.

In order to study the conflict characteristics of vehicles in the merging area on the highway, the conflict risk of vehicles in the merging area on the highway is reduced. This paper is based on continuous microscopic trajectory information obtained by LiDAR. An active traffic safety analysis method is proposed and the existing unconstrained conflict discrimination method is improved. Focus on the proximity of two or more road users in time or space and the likelihood of accidents. This method can effectively overcome the shortcomings of inaccurate records and incomplete analysis of traditional accident statistical methods. Further expand the practical application scope of this type of trajectory, while verifying the accuracy of trajectory tracking, calculate traffic conflicts from trajectory data, and analyze the potential collision risk of each target in the trajectory, so as to provide technical and data support for reducing traffic accidents.

To avoid the problem of low AV (Autonomous Vehicle) lane utilization due to low demand for autonomous vehicles. A new type of dedicated AVT (Autonomous Vehicle Toll) lane is considered, where Human-driven vehicles are allowed to pay to enter the AVT lane and autonomous vehicles are allowed to enter the AVT lane free of charge. The optimal allocation of dedicated AV lanes and AVT lanes in a traffic network with mixed autonomous and human-driven vehicle flows is investigated, saving the total travel cost of the transport system. A bi-level programming model with a bi-level recursive structure is developed for the optimal deployment of lanes considering mixed traffic outflow conditions of autonomous and human-driven vehicles. The upper-level model is a mathematical planning model for decision makers to optimize the total travel cost of the system using decision variables, and the lower-level model is a mixed traffic equilibrium assignment model when autonomous vehicles follow the UE travel principle and human-driven vehicles follow the SUE travel principle to select their paths. The model is solved using KT complementary relaxation conditions and sensitivity analysis is performed on the key parameters in the model. The results show that the optimal deployment of AV/AVT lanes is related to the perception factor and the market penetration of AVs. The optimal network design solution for AV/AVT lane mixed deployment is not only effective in improving traffic efficiency, but also reduces the overall system cost of the traffic network and road traffic congestion.

The confluence area of highways is a high-risk area for traffic accidents and congestion. By analysing vehicle trajectory data, the driving style and bad driving behaviour of drivers are identified, to take timely measures to prevent traffic accidents, alleviate traffic congestion, and improve traffic safety. This article is based on vehicle trajectory data to construct driving behaviour feature parameters. PCA is used to reduce the dimensionality of the feature parameter matrix to remove the linear correlation between the feature parameters. The principal component contribution rate reaches 96%, and the reduced feature matrix is used as input to the K-Means++ algorithm for clustering. The clustering center K is determined using the contour coefficient method. Compare the results with the K-Means results. According to literature experimental analysis, driving styles in confluence areas can be divided into three categories: aggressive, steady, and calm. The experimental results are scientifically reliable and provide important references for traffic management, road design, driver education, and intelligent transportation systems.

The protective effect of highway guardrails on vehicles is closely related to collision speed and collision angle. With the development of road traffic, it is possible to construct super high-speed highways. When super high-speed vehicles collide with guardrails at different angles, the protective effect of guardrails will vary. In order to analyze the guardrail protection effect during vehicle collision with guardrails, a numerical simulation method was used to model and analyze the guardrail collision process of the super high-speed car, taking into account the impact of different collision angles of the car. The main conclusions are as follows:(1) During the collision of the super high-speed car with guardrails at different angles, the cars did not cross, break through, or ride over the guardrails. The car's wheel tracks were all within the range of the guided exit frame, and the guiding and buffering functions of the guardrail structure met the requirements. As the collision angle between the vehicle and the guardrail continues to increase, the vehicle's wheel track becomes closer to the long side of the guiding exit frame away from the guardrail. As the collision angle increases, the guiding function of the guardrail on the vehicle decreases. (2) As the collision angle between the vehicle and the guardrail increases, the buffering effect of the guardrail on the vehicle decreases, and the risk of passenger injury increases. When the collision angle between the vehicle and the guardrail is 30°, the collision speed and acceleration of the passengers exceed the standard requirements. (3) As the collision angle of vehicles increases, the collision force between cars and guardrails increases, and the collision load acting on the guardrail structure increases. Under larger loads, the deformation of the guardrail structure increases, and the risk of car collision with guardrails increases.

The characteristics and laws of the response of the vehicle to the cable structure of the cable-stayed bridge under different driving conditions are widely concerned by the engineering community. In this paper, the finite element model of cable-stayed bridge is established by using the nonlinear finite element software Abaqus, and the stress variation characteristics and laws of cable are simulated when the vehicle driving conditions on the bridge are different, so as to provide the basis for the design improvement of cable under moving load.

The key to traction control in locomotive systems is adjusting the axle traction torque through adhesion control. In traditional adhesion control systems, due to the inability to detect the peak adhesion coefficient of the rail surface in realtime, locomotives are prone to spinning and sliding phenomena under different rail surface conditions. This reduces the utilization of adhesion and can even lead to irreversible damage such as tread detachment and rail wear. Therefore, this paper proposes a machine learning-based rail surface recognition method to identify the rail surface condition online and set the peak adhesion coefficient for different rail surfaces. The output torque of the traction motor is dynamically adjusted based on the set reference peak adhesion point to fully utilize the current adhesion force between the rail surface and wheels. A locomotive multi-axis adhesion control model is established to validate the effectiveness of the rail surface recognition algorithm. The results show that the improved adhesion control method effectively prevents wheel spinning, improves the stability of locomotive operation, and reduces the loss of adhesion utilization.

China has unveiled guidelines on developing a comprehensive transport network, with the aim of creating a modern, highquality and comprehensive national transport system. Amid efforts to optimize the country's waterway infrastructure layout, the total length of the high-grade waterways is projected to reach approximately 25,000 km by 2035. It is expected that the construction of high-grade waterways will experience rapid growth from the 14th Five Year Plan period to the mid-century. In order to guide and standardize the construction of smart waterways, this study analyzes the connotation of smart waterways. Furthermore, it elucidates the technical framework of smart waterways and establishes an evaluation index system to assess the effectiveness of their construction. This found paves the way for the integrated and coordinated development of smart waterways.

In order to improve the construction quality of swivel bridge, the measurement control method of swivel bridge across railway is studied. According to the actual situation of the project, the construction control network of the swivel bridge of plane and elevation railway is constructed. According to the change of the position of control points, the first level position value of different measuring control areas can be adjusted in time. GPS technology is introduced to control the precision of construction survey by adjusting the control network. The experimental results show that the method can effectively improve the measurement accuracy and guarantee the construction quality of railway swivel bridge

In recent years, China's highway network has developed rapidly, and the demand for highway infrastructure maintenance has increased rapidly. Most of the existing research on highway maintenance focuses on optimizing the maintenance strategies under the cost constraint from the perspective of the highway agency, but ignores that the actual highway maintenance activities exist complex conflicts of interest. In the highway maintenance industry, the conflicts of interest between the highway agency and the maintenance service providers are inevitable, but a game equilibrium strategy can reduce the conflicts of interest between them. This paper starts from the dynamic game perspective of the highway agency and the maintenance service providers to establish game-based model, where the highway agency is deemed to be the leader determining the maintenance mileage and the maintenance service provider is deemed to be the follower providing appropriate maintenance services. The dynamic game-based model aiming at finding the optimal mileage is proposed to handle the conflicts between highway agency and maintenance service provider. To solve the model, the Sub-game Perfect Nash Equilibrium for this dynamic game problem is calculated to find the maximum maintenance mileage for the highway agency. A 128km highway maintenance project in Zhejiang Province is used to verify the game-based model, and the game analysis result of the maximum maintenance mileage interval is (41.71km, 45.68km).

There are many problems in traditional survey methods for existing railway lines, such as single point measurement, low efficiency, high safety hazards, and high labor costs. In order to overcome the above shortcomings, this paper proposes a collection and extraction method of Railway track alignment based on the ground 3D laser scanning point cloud data: under the control point stitching principle, the original point cloud data is collected in a certain railway section, and the Correspondence principle is used to "one button" splice the original point cloud data; The original point cloud is selected, the elevation is extracted, the mixed points are filtered and the point cloud is reduced to obtain the required Railway track point cloud data; The processed track point cloud is piecewise fitted with the least square method to obtain the existing railway track alignment.

In order to understand the interval design of expressway comprehensive energy system, a research on interval design of expressway comprehensive energy system based on multi-objective is put forward. From the perspective of joint planning and operation, taking the planning result as the direction and the operation result as the target calculation basis, this paper establishes the relationship between planning and operation from the perspectives of annualized cost of equipment investment in the planning process, investment time value, energy use cost and operation and maintenance cost in the operation process, and realizes the integrated planning and operation optimization model of comprehensive energy system. Secondly, the optimization process, optimization goal and solving algorithm of the model are put forward. Through examples, it is proved that the load interval of electric vehicle is basically consistent with the trend of electric load. If it is not adjusted, it is easy to cause the phenomenon of "peak plus peak" of electric load. Wind power has a large output at night, and it is easy to produce wind curtailment, so we can focus on how to absorb the excess wind power in these periods.Finally, according to the actual construction needs of users, the construction scheme of comprehensive energy system under fixed investment constraints is put forward, which provides theoretical basis for solving the planning problem of comprehensive energy system.

The development of transportation has shifted from construction to the state of maintenance, and reasonable and efficient pavement performance prediction models play an important role in developing reasonable pavement maintenance strategies, ensuring the efficient operation of pavement and extending the service life. At present, the research on pavement performance prediction mostly focuses on a non-comprehensive pavement performance evaluation indicator, and when the influence factors increase, the predictive ability will also be affected. Therefore, based on deep neural networks, OPI is selected as the pavement performance evaluation index, and the pavement service time, traffic load, pavement asphalt rutting condition and pavement asphalt cracking condition are selected as the influence factors. Next, the pavement performance deterioration model is established, and the established model is compared with the traditional regression model. After comparison, it is found that the prediction accuracy of the deterioration model reaches 83.25%, which is 17.7% higher than the traditional regression model. Besides, the prediction accuracy of the vast majority of test sets and train sets exceeds 80%. According to the research in this paper, the pavement performance deterioration model based on deep neural networks can better simulate the pavement performance deterioration process, and can still predict pavement performance well under multiple influence factors.

Traveler classification is of practical significance for the operation and management of highways. In order to refine the analysis of vehicle travel distribution characteristics, promote effective measures to use data for traffic theory research, and provide better management for highways, this paper proposes a classification model using highway toll data in Shandong Province. This paper combines big data to analyze the spatial and temporal distribution characteristics of travel users. Seven indicators were constructed from four dimensions: traveler's perspective, travel time, travel intensity, and travel space. Then we use K-means++ algorithm to cluster the full sample of travelers. Travelers are divided into five categories, which are commuter travelers, life travelers, work travelers, sporadic travelers and other travelers. The results show that the proposed model can effectively classify highway travelers, and the distribution of travelers' characteristics varies significantly among categories, which can provide reasonable suggestions for highway management.

With the rapid development of automatic driving technology, lv3 level vehicles have now landed, which indicates that the environment of automatic driving and manual driving environment will exist for a long time. The setting of highway automatic driving lanes can provide a more efficient and safer driving environment for autonomous tvehicles. In this paper, by constructing a simulation model and combining the designed simulation scheme, research is carried out on a two-way six-lane highway to explore the conditions for the setting of automatic driving lanes and analyze the results with the speed and the average delay as the indexes. The conclusion is drawn: When one dedicated lane is set up, the proportion of automatic driving should be no less than 60% when the traffic volume is small; when the traffic volume is large, the proportion of automatic driving should be no less than 20%. When two dedicated lanes are set up, it is necessary to ensure that the proportion of automated driving is not less than 40% when the traffic volume is small;When the traffic volume is large, the proportion of automatic driving is not less than 20%. And it is expected to provide a certain reference for the subsequent research on the setting of automatic driving lane.

Meteorological impacts on highways play a crucial role in transportation safety, efficiency and planning. Heavy rainfall, snowstorms, fog and other extreme weather events can disrupt road conditions, visibility and vehicular movement resulting in traffic congestion, accidents and travel delays. Meteorological data, collected through weather stations, sensors and satellite observations, are essential for monitoring and predicting weather-related impacts on highways. There are so many different extreme weather events, it is hard to recognize the intricate relationship between meteorological conditions and highway accidents. However, the relationship is vital for enhancing safety, reducing travel disruptions and ensuring efficient transportation systems. The paper analyzes accident data from 2020 to 2022 alongside meteorological data to identify patterns and correlations between specific weather conditions and accident rates. The results show that approximately 40% accidents occur during extreme weather condition. The heavy rain, snowstorms, fog and ice can significantly impact on highway road conditions leading to an increased likehood of accidents. Based on the analysis, intelligent traffic system and automobiles make informed decisions ragarding vehicle operation and road safety. The proposal from the analysis has been applied to Hebei-Beijing-Harbin Expressway's pile-by-pile meteorological monitoring project.

The prediction of skid resistance of asphalt pavement plays a pivotal role in formulating maintenance plans and determining maintenance schemes. At present, the typical intelligent algorithms such as the neural network and the genetic algorithm have seen extensive applications in the evaluation and prediction of pavement performance. The combination forecasting model can leverage the complementary advantages of the two, thereby enhancing the reliability of prediction. As a case study, this paper focuses on the prediction of pavement skid resistance for an expressway in Chongqing. The research establishes a pavement skid resistance forecasting model using a genetic neural network and compares it with the single neural network, genetic algorithm, and regression models. Through this comparative analysis, the study validates the applicability and reliability of the genetic neural network approach for predicting asphalt pavement skid resistance. The results demonstrate that the regression model exhibits a limited fitting degree for highly nonlinear problems, leading to noticeably lower prediction accuracy than the genetic algorithm or neural network algorithm. In contrast, the combination forecasting model significantly enhances prediction accuracy in comparison to a single neural network model or genetic algorithm. Notwithstanding, it is worth noting that the operational efficiency of the combination forecasting model is inferior to that of a single neural network model. Consequently, the genetic neural network combination forecasting model proves more suitable for pavement skid resistance prediction, and nevertheless, there is room for further improvement in the operational efficiency of the model.

To construct an automatic discrimination method for the causes of pavement diseases, the typical characteristics of different types of asphalt pavement diseases of the Inner Ring Expressway in Chongqing, which was taken as an engineering example, were analyzed, and the feasibility of data dimension reduction analysis was determined based on the correlation characteristics of different types of damage. Then, numerous state information data were subjected to dimension reduction through the principal component analysis (PCA), followed by the automatic cause analysis of pavement diseases using the random forest algorithm. The results show that the cause conclusions acquired through machine learning model training basically accord with the actual field survey conclusions. Thus, it can be deemed that the intelligent discrimination method based on machine learning is reliable, to some extent, for the cause analysis of pavement diseases and can serve as an automatic discrimination method for the follow-up development of an intelligent maintenance decision system.

Accurate toll prediction is a concern for expressway operating companies. However, due to factors such as time period, week sequence, and weather, the toll on expressways often exhibits extreme imbalance, which also makes accurate prediction of toll very difficult. In response to this issue, this paper proposes a segmented expressway toll prediction model. Using the toll anomaly detection algorithm, the training samples are divided into multiple sub sample sets, and the toll is mapped to the most relevant features. Then, the models are trained separately based on the partitioned dataset. The experiment shows that the regional central city expressway toll model established based on the method presented in this paper has a much better performance than similar traditional models.

At present, with the rapid development of the transportation industry, the investment scale of the electromechanical project of the expressway is large, the construction mileage is long, and the design task is scattered, which makes the technical parameters of electromechanical equipment vary greatly, and there are some differences in the performance parameters of electromechanical equipment in different climatic environment. Therefore, building a method system of electromechanical key equipment selection based on environmental characteristics is not only a realistic requirement but also a formal requirement. Based on this, firstly, price, quantity, equipment load, and other indicators are selected as considerations to classify the key mechanical and electrical equipment of expressways in Xinjiang by ABC. Secondly, the parameters of the core equipment are recommended. In addition, the influencing factors of electromechanical equipment parameters are analyzed by the analytic hierarchy process, which provides a basis for the selection of electromechanical engineering equipment and the process of parameter determination.

After an accident occurs, traffic risk propagates upstream along the roadway. Within the spatiotemporal impact area of the accident, traffic flow accident risk experiences fluctuations with uncertainty. Accurately describing the process of traffic accident risk propagation on the road and quantitatively analyzing it will provide a solid theoretical basis for traffic risk management measures. Therefore, this study focuses on the risk propagation mechanism and prediction methods after traffic accidents in highway diversion areas. Firstly, this paper introduces the concept and indicators of risk propagation. Secondly, through simulation experiments, the driving scenarios in the diversion area are simulated with different traffic volumes, and the regularities of risk propagation are studied by observing changes in risk indicators. Finally, using a simulated dataset as input, a GRU (Gated Recurrent Unit)-based accident risk prediction model is constructed. Experimental results indicate that in predicting accident risk, GRU exhibits higher prediction accuracy compared to LSTM (Long Short-Term Memory).

This paper proposes a multi-level comprehensive fuzzy evaluation algorithm for highway route schemes that considers operating speed and association rules. By mining multiple evaluation factors of the route plan through association rules, an evaluation index system is constructed. In order to improve the efficiency of the algorithm, the evaluation factor data is preprocessed. According to the fuzzy decision algorithm, the route schemes are sorted and multi-level comprehensive fuzzy evaluation is achieved. The experimental results show that the proposed multi-level comprehensive fuzzy evaluation algorithm can obtain accurate and efficient evaluation of route schemes while considering running speed and association rules. This algorithm has practical application value in fields such as transportation and tourism planning, and can provide users with personalized travel experiences.

Model coding is crucial for the integration and effective transmission of digital highway information models at various stages throughout their lifecycle, enabling integrated model management and analytical applications. By studying the construction methods of domestic and foreign highway engineering information model identifiers and analyzing the advantages, disadvantages, and current applications of engineering project decomposition structures, spatial + classification coding, considering different needs such as project construction and operation management, engineering measurement, archives management, and big data analysis, a "five-segment" digital highway model classification coding framework based on the ISO 12006-2 is proposed. Subsequently, an automated coding approach based on parametric modeling and rule engines is introduced, significantly enhancing the efficiency and accuracy of coding work. The practical application of digital highway model identification coding in highway engineering information systems demonstrates its ability to establish effective connections between information models and smart construction management business data, thereby providing support for the sharing and transmission of data throughout the entire lifecycle of highway projects.

Based on the establishment of node centrality, this study comprehensively considers the optimality of the highway network layout optimization model and develops a highway network layout optimization approach based on comprehensive centrality. Traditional heuristic algorithms often face challenges of falling into local optima and low solution accuracy when solving layout models. Therefore, to enhance the search and developmental capabilities of the Whale Optimization Algorithm (WOA), this research introduces a non-linear convergence factor. A case study on the optimization of the highway network layout in the Xi'an metropolitan area validates the rationality of the progressive and refined highway network layout optimization theory and methodology. The comprehensive centrality model for nodes comprehensively reflects the impact of various transportation modes around nodes on their development. This calculation, which avoids the need to determine weights, is more comprehensive and objective. The road network layout combinatorial optimization model developed in this study balances mathematical optimization principles with the rationality of layout configuration. It effectively optimizes the highway network layout to meet the requirements of an increasingly mature highway network layout, thus maximizing the scale benefits and overall efficiency of the network. This provides fresh perspectives and methodologies for highway network layout and enhances the understanding of highway network planning.

Freeway construction conditions are complex and require a system that can take these complexities into consideration in the modeling phase so as to optimizing time and cost savings. For this reason, international freeway digital system is established through BIM building information modelling, which integrates various information in the system and can reflect the details of each stage of the freeway project. In this study, which is based on the international freeway project, the modelling method of the freeway BIM digital system is formulated, and the application of BIM technology in the design and construction of international freeways is explored, which is effective in the construction of freeways. A collaborative management platform based on the BIM model was established to promote the development of intelligent construction and management of freeways.

Rainfall weather can easily lead to problems such as wet and slippery pavement, reduced friction coefficient and low visibility, which directly affect the safety of highway traffic. With the continuous increase of highway mileage and travel vehicles, the probability of accidents in rainy weather is increasing. In order to reduce the occurrence of road traffic accidents under rainfall weather, improve the accuracy and timeliness of rainfall disaster warning and management, pro-vide a more accurate and effective basis for meteorological disaster risk classification, and carry out research on refined rainfall disaster risk classification scheme. This paper summarizes the existing rainfall disaster risk classification standards, constructs the meteorological disaster index model, and proposes a lean rainfall disaster risk classification scheme from three aspects : possibility, severity and sensitivity.

This paper carried out risk assessment of the immersed tube tunnel engineering using the Bayesian networks. Firstly, based on expert survey data and records of construction risks in immersed tube tunnel engineering the fault tree is transformed into a systematic Bayesian network. Next, leveraging the computational advantages of the Bayesian method, the probability of occurrence of risk events in the immersed tube tunnel is linearly inferred, identifying the risk factors that require particular attention. Finally, it is proposed that when any non-root node event fails in the Bayesian network, the root node events should be controlled in order of decreasing posterior probability. This control method clarifies the approximate scope and quantity of key control targets, to some extent, conserving limited human, material, and financial resources. The evaluation results can provide a scientific basis for the subsequent risk control.

Highway freight transport emissions provide a high contribution to carbon emissions in the transport industry, and analysing the driving factors is conducive to building a low-carbon transport structure. This paper explores the indicators of economic dynamic series level, traffic development and energy consumption structure that affect the carbon emissions of the transport industry by using the pathway analysis, and analyses the dependence of carbon emissions on the truck model through visualisation. The results show that the direct throughput coefficients of GDP of the secondary industry, freight turnover, and the structure of truck models are 0.590, 0.317, and 0.204, respectively, indicating that the economic growth has a positive pulling effect on the carbon emissions of the transport industry, and the reliance on the other two types of indicators is small. The indirect throughput coefficients were 0.358, 0.619 and 0.613, indicating that transport development and model structure are highly dependent on the economy, and that adjusting the model structure with the help of economic development can effectively promote the implementation of emission reduction measures.

An investigation was conducted to predict ship trajectories and discern their navigational intent in intersecting waters. The study involved experiments utilizing AIS data near the Baoshan District, Shanghai, and the development of the RBFNN+KNN model. The data preprocessing steps encompassed: 1) cleaning abnormal trajectory points and routes; 2) employing the LOF algorithm to filter trajectories with significant outliers; 3) clustering ship trajectories. Two clusters, representing routes in the straight-ahead channel and the right-turn channel, were selected as the sample dataset. The RBF neural network was employed to forecast ships' trajectories in future time periods, and the KNN classification algorithm was integrated to determine the vessels' sailing intentions. The experimental outcomes demonstrated that the proposed method achieved 99.88% accuracy in navigational intent recognition for the training set and 99.18% for the test set. These results indicate the model's superior ability to predict ships' navigational intent.

Intelligent freeways are a type of freeway construction model based on information technology and intelligent technology. In order to facilitate research on intelligent freeways, this paper proposes SimIWay, an integrated simulation software platform for intelligent freeways based on SUMO and Unity3D. SimIWay consists of three parts: road network and traffic flow generation, integrated simulation tools, and application scenarios. The combination of these three partsforms a full-stack simulation platform, including functions such as traffic flow generation, freeway network construction, traffic simulation, traffic control, traffic data analysis and visualization. It supports importing real road networks and traffic flows from the real world, and can also customize virtual road networks and traffic flows according to requirements. The simulation tool consists of SUMO, Unity3D, and driving simulators, providing a realistic freeway simulation environment. The platform is highly modular and can simulate a variety of intelligent freeway application scenarios. We chose the example of variable speed limits (VSL) on freeways. The VSL control uses an extended Cell Transmission Model (CTM), which aims to improve the efficiency of freeway operations by formulating a single objective function and using a genetic algorithm for solving the speed limit values. The simulation of this scenario demonstrates the platform's ability to research intelligent freeways.

Under the background of "Internet of Things + big data", the accident analysis and safety prevention of long and steep sections of highway have attracted the attention of academic circles. Reasonable analysis of safety influencing factors can contribute to ensuring traffic safety of highway projects and improve prevention technology. Therefore, based on multi-day and multi-source traffic accident data in Guizhou province, this paper constructs a Multinomial logit(MNL) model considering multiple accident causes. Firstly, data related to accident causes are screened out from multi-source data, and these data are divided into two categories of traffic accident characteristics and road environment characteristics. Secondly, based on the multi-day survey data, the MNL model was established by using python to conduct regression analysis on the relationship between the accident influencing factors and the accident severity under the general road section and the long and steep road section respectively, and the two were analyzed and compared based on the significance factors. Finally, the accident data of long and steep longitudinal slope sections are selected to do univariate regression analysis of the annual accident number and each influencing factor. An empirical experiment was carried out in Guizhou Province. The results show that the accident type, vehicle type and accident location are the obvious differences between the long and steep slope road accidents and the general road accidents. The coupling effect of slope length and slope, traffic flow and truck proportion have great influence on the number of annual accidents. This study can provide theoretical support for the improvement of highway accident prevention technology in mountainous areas

In order to better describe the flight support vehicle scheduling system, a fuzzy colored Petri net (FCPN) model is proposed, and its formal definition and specific inference algorithm are listed. A dispatching model with active strategies is developed using FCPN. An application example yields that the FCPN model significantly reduces the model complexity and is able to more accurately describe the vehicle scheduling system.

To solve the problems of low detection efficiency, high safety risk and production cost of urban pipe culverts and drainage outlets, a suspended intelligent detection system has been developed. The system adopts self-developed RGBD camera and amphibious suspended carrier, and develops 3D reconstruction technology of underground space, which can realize the output of panoramic 3D models. Test results show that the robot has strong adaptability to the underground environment, and the modeling error and positioning accuracy can meet the engineering requirements. The system overcomes the limitation that detectors installed on traditional carriers cannot fully adapt to the complex environment inside the pipe culvert, and has been applied in several water environment and utility tunnel projects.

Taking the bridge track installed in orchard rail transportation as the research object, based on the SIMP variable density method topology optimization theory, the main components of the bridge track main beam are topologically optimized and designed. First, ANSYS is used to establish a finite element model of the bridge type track, and combined with the engineering application situation and the connection characteristics of the main beam of the bridge type track, lightweight design is carried out. Second, the static analysis and modal analysis of the lightweight bridge track are carried out, and the stress, deformation, mass and natural frequency of the whole structure of the main beam before and after the lightweight are compared under the actual working conditions. The results show that even with a reduction of 18.25% in the mass of the main beam, the bridge track still meets the actual operating requirements, and the first natural frequency of the lightweight bridge track is 23.614Hz, which is greater than its maximum operating frequency, and resonance phenomenon will not occur.

In the tunnel traffic environment, good lighting conditions are essential not only for ensuring safe driving but also for improving tunnel passage efficiency. Highway tunnel lighting differs from general road lighting as the luminaires are required to operate under all weather conditions. However, due to various factors, some lighting fixtures may experience light decay, which affects the lighting quality of the highway tunnel[1]. By drawing the tunnel road surface illuminance curve, the operational status of the tunnel lighting system can be understood, enabling an accurate assessment of the lighting quality. This study utilizes the inverse distance weighting (IDW) interpolation method and the kriging interpolation method to perform plane interpolation of tunnel road surface illuminance, resulting in the generation of the tunnel road surface illuminance curve. It is observed that the kriging interpolation method using the exponential function as the theoretical model for the semivariogram produces the best interpolation results.

With the development of information technology, BIM technology has been widely used in the construction industry. In comparison to BIM3D technology, BIM5D technology integrates cost information and progress information. Through the dynamic simulation of the construction management process, the efficiency of the construction management process is improved. This paper takes residential building as the research object, uses BIM5D technology to study the dynamic management of its building construction, and combines the Internet plus technology to conduct unified management of building construction. The results show that BIM5D technology brings together all stages and links in the construction process to form an intelligent shared platform, which effectively improves the efficiency and accuracy of construction projects.

In the process of integrated processing of vehicle road cooperative signals, due to the low accuracy of the analysis results of road traffic status, the specific effect of using it to manage road traffic is not ideal. For this reason, this paper proposes the research of vehicle road cooperative signal active integration algorithm based on deep reinforcement learning. Taking full account of the shortcomings of the reinforcement learning algorithm in terms of convergence, we introduced deep reinforcement learning, added a new target network to the original reinforcement learning network, combined with the road traffic state transition matrix, and calculated the traffic state of the road environment by updating and iterating. In the active integration stage of vehicle road cooperative signals, based on the boosting algorithm, the differences shown in the analysis results of road traffic status are used to construct corresponding differences between integrators, and the integrated processing of them is realized on the basis of differentiated road traffic status signal weights. In the test results, under the design algorithm, the average travel time of road traffic is only 185s, the average travel speed reaches 5.97m/s, and the average queue length is only 6.52 vehicles, which has good application effect.

To obtain object micro traffic information, LiDAR plays an important role in intelligent perception. Many background filtering methods based on LiDAR can detect the target, but the data process is not simple enough and data information will be lost easily. In this work, from the principle of original point cloud data capture, an advanced background filtering algorithm based on PCAP (Packet Capture) file is proposed. Firstly, a data extraction algorithm for PCAP files is designed. Then, one non-target frame is chosen as the background frame. The distance difference method is used to filter the background under the spherical coordinate system with the relative horizontal angle and vertical angle as discrete coordinates. The algorithm is improved twice by reconstructing the background frame based on the hierarchical clustering method and changing the distance difference threshold. Finally, the effectiveness of the three algorithms is verified.

To improve the level of over-under excavation detection in construction process of mining tunnels, scholars demostic and abroad have combined 3D laser scanning with other information technology to achieve certain research results. However, the existing research results mainly use sectional method, without fully taking advantage of 3D technology. To address these issues, this paper uses algorithms such as registration, cutting, denoising, and downsampling for pre-processing of laser scanning point clouds. Then, point clouds are transformed into spatial meshes through cluster and BPA algorithms. With the help of triangular mesh projection and prism volume integration, 3D over-under excavation condition is calculated. Through application verification in actual engineering projects, it is achieved that the over-under excavation detection process during construction phase for mining tunnels are automated and intelligent.

The increasing number of vehicles has led to issues such as traffic congestion, air pollution, and energy consumption. Intelligent connected autonomous driving (AD) offers a solution to these transportation challenges, but it requires cooperation among dedicated AD road networks, connected cloud-based control and dispatch systems, and intelligent vehicle sensing. To harness the potential of AD, it is crucial to consider the requirements of autonomous vehicles (AVs) when planning road networks. Well-designed road networks can enhance travel experiences, reduce energy consumption, and promote urban sustainability. In order to meet the needs of AD while ensuring the operational efficiency of urban transportation networks, this paper introduces a road network evaluation framework that encompasses both static and dynamic assessment criteria, following a detailed exploration and analysis of various road network generation methods. This evaluation framework addresses the deficiency in existing road network generation methods, which lack dynamic assessment of traffic networks. To validate the proposed approach, we have chosen a road network generation method based on conditional generative adversarial network (CGAN) as an application example. Through carefully designed experiments, we confirm the feasibility and effectiveness of the proposed method for generating and evaluating road networks tailored for AD scenarios.

In order to improve the uniformity of wheel track transverse distribution when vehicles drive on asphalt pavement, delay the development of the asphalt pavement disease, and effectively avoid the waste of asphalt pavement design and maintenance resources, the relationship between vehicle wheel track distribution and road surface disease distribution is analyzed by the accurate identification method of traffic axle load dividing lane based on surveillance video, and then the influence of lane markings on wheel track distribution and asphalt pavement diseases was studied. Results show that: (1) The proportion and distribution of different vehicle types are accurately identified by SVM classifier based on object detection algorithm; (2) There was a high consistency between the distribution of asphalt pavement load-type diseases and the lateral distribution of vehicle wheel tracks; (3) The vehicle trajectory and pavement disease distribution on the asphalt pavement distributed more evenly if lane markings were swung and shifted. Based on above results, the swing and shifting method of lane markings, is proposed to achieve the redistribution of asphalt pavement wheel tracks by "swing and shifting" of lane markings. By using this method, the effective utilization rate of asphalt pavement could be improved, and asphalt pavement diseases could be improved, thus the service life of asphalt pavement could be extended. It provides a new idea for future maintenance of the asphalt pavement.

As China's economy continues to develop and people's living standards continue to improve, travel and tourism has become an important way for the public to relax and relieve the stress of life, and along with the introduction of smart tourism and the rapid development of information technology, new forms of tourism, mainly self-guided tours, have begun to emerge and flourish. This paper examines the optimisation of tourism routes to maximise the needs of contemporary people, taking into account the new contemporary fashions of routes and travel styles throughout history. In addition to planning viable and quality routes, people are provided with the best possible service and experience. This paper is guided by a new combination of research based on the context of big data, combining an in-depth study of computer modelling algorithms based on the creation of urban tourism itinerary chains.

In the social force model, interpersonal forces are manifested as the space between moving subjects, which serves as a reference for the range of forces between the target and other subjects. Although numerous studies have investigated the spatial requirements of road users, validating these findings using large, real-world datasets remains challenging. This study focuses on "safe spaces" for pedestrians and e-bicyclists during street crossings, considering both psychological and safety needs. Drone-captured video data from three major intersections in Xi'an, China, is utilized to examine these safe spaces. The findings indicate that pedestrians maintain semi-circular safe spaces while e-bicyclists maintain semi-elliptical safe spaces when crossing the street. Moreover, as the speed increases, the safe space expands. However, the safe space does not increase with speed indefinitely. The size of the safe space fluctuates in a certain range when the speed increases to a certain level. It is worth mentioning that e-bicyclist require greater safety distances in the forward direction than in other directions and are more influenced by speed.

In order to improve the elasticity of urban rail transit (URT) network, considering action and recovery process of disruption events, a risk assessment method for URT network based on robustness, vulnerability and elasticity is proposed. By quantifying the overall impact of disturbance on the performance of the URT network, a recovery model is established to maximize the elasticity index of the URT network and an adaptive genetic algorithm is used to solve it. Taking Xi'an URT network as an example, two disturbance scenarios are examined: a random attack and a deliberate attack. The recovery effect of the target recovery strategy, the random recovery strategy and the preferred recovery strategy on the elasticity of the URT network after the disturbance occurs is comparatively analyzed. The results show that the recovery effect of the targeted recovery strategy on the elasticity of the URT network is the best, followed by the preferred recovery strategy. The conclusions of this paper can provide strategic support for the recovery of the URT network after disturbance.

To determine the risk levels of urban signal-controlled intersection lanes, we propose two comprehensive risk assessment sets that incorporate time proximity, spatial proximity, and velocity-based indicators. Using actual trajectory data of intersection users, we analyse the interactions between three common vulnerable road users and motor vehicles. Bayesian inference algorithm is employed to evaluate lane risk levels. Additionally, XGBoost, random algorithm, and SVM algorithm with a Gaussian kernel are utilized to predict intersection lane risk. The importance and correlation of risk factors are explored using SHAP value theory. Results indicate that while extremely dangerous collisions are less likely due to signal control, moderate risk remains prevalent. The XGBoost model outperforms the other methods with an accuracy rate of 91.51% and precision rate of 94.17%. Among the analysed interaction risk factors, the speed and acceleration of road users have the greatest impact on lane risk levels. In low traffic flow conditions, higher speeds of motorcycles and cars increase collision risks. Lanes three and four exhibit higher probabilities of hazardous interactions between pedestrians and cars, resulting in more severe pedestrian injuries. Additionally, motorcycles at lower speeds are more prone to risky interactions with pedestrians.

Around the research questions of this paper, the main theories involved in this paper are analysed, including the relevant concepts of smart cities, systems science theory, urban systems theory, urban planning theory and other basic theories, so as to have a clearer understanding of the basic theories of cities, complex systems, urban systems and urban planning, in order to clarify the theoretical inheritance and expansion relationship between this paper and the current research results, and to lay the theoretical foundation for the following analysis of smart cities, smart city systems and the construction of smart city system models

The industrialization of electric vehicles has brought with it a large number of used batteries. In order to support the development of electric vehicles, it is necessary to recycle waste batteries in an efficient way. This paper investigates the path planning problem in the recycling process of used batteries for electric vehicles. The recycling path is required to be optimized to minimize the total cost. In contrast to existing research, a multi-level recycling network for used batteries in electric vehicles is designed taking into account practical scenarios, while the balance of vehicle working distances is introduced into the objective function. In the model construction, the workflow of the recycling network is designed in a distributed manner and the variance weighting of the vehicle working distance is introduced into the objective function. The study combines iterative local search algorithm with variable neighborhood search algorithm to design a multi-objective iterative local optimization algorithm. The improved model is proven with the algorithm to solve practical problems and can be used for future research.

To explore the effect of transportation networks on the formation of passenger connections in urban agglomerations, this paper takes 90 counties in Guanzhong Plain urban agglomerations as the object of analysis, and extracts passenger transport connections in different periods of time by using mobile phone signaling data. GIS analysis, distance attenuation function and geographically weighted regression model were used to explore the heterogeneity of driving effects of transport network on passenger transport connections during the May Day holiday, working day and weekend. The results show that the travel of urban agglomerations is mainly concentrated in the central region of urban agglomerations, and the contact range of districts and counties is wider during the May Day holiday, and the peak value of weekday travel is higher. The passenger flow in urban agglomeration decreases with the increase of time and space distance, and the fitting degree of the power-law attenuation function is better than that of the exponential attenuation function. The ranking of the power-law attenuation coefficients of the three periods is May Day holiday < working day < weekend. The density of the main roads and the total number of bullet train connections have a positive driving effect on the total number of passenger transport connections, while the total number of common railway connections shows a suppressive effect on the total number of passenger transport connections. The driving effect of all factors on the total number of passenger transport connections has obvious heterogeneity in both time and space.

In order to further analyze the potential causes and influencing factors of Marine traffic accidents in Yangshan Port waters, 87 Marine traffic accidents above grade occurred in Yangshan Port waters from 2015 to 2021 were taken as data samples. A comprehensive Marine accident Analysis framework HFACS-BN (Human Factors Analysis and Classification System (HFACS) and Bayesian Network (BN)) is established. In addition, the contribution degree and correlation of 4 primary nodes and 15 secondary nodes are quantified by the method of grey correlation degree analysis, and the possibility and severity of accidents are further analyzed to evaluate the potential impact of various factors on the occurrence of Marine traffic accidents. Finally, the reliability of the model is checked by cross-validation method. The results show that "the premise of unsafe behavior" is the key cause of Marine accidents and should be widely concerned. This study has important theoretical and practical significance for preventing and managing potential Marine accident risks.

In order to investigate the differences in causes of motor vehicle and non-motor vehicle accidents between day and night, reduce the losses caused by accidents, and establish the Ant Colony Algorithm (ACA) to mine the association rules between different severity levels of motor-nonmotor traffic accidents and various factors. Firstly, the association rule generation will be transformed into the Traveling Salesman Problem, and then the Ant Colony Algorithm will be utilized for solving it, aiming to enhance the quality of the rules. Secondly, directional constraints will be applied to the rules to reduce the generation of redundant rules. Lastly, an analysis was conducted on 2178 motor-nonmotor accident data. The results indicate that the proposed ACA algorithm can significantly reduce the number of invalid rules and accurately identify the correlation between potential accident factors and severity levels. Compared to daytime, there is a greater tendency for higher severity level accidents to occur at night due to environmental and non-motor vehicle driver factors. Furthermore, there are significant differences in the correlation between factors such as rain, fog, snow, driver alcohol consumption, and speeding, with the severity of accidents occurring during the night and daytime.

In view of the characteristics of the spatiotemporal imbalance of passenger flow in urban rail transit sections, the optimal allocation of resources is realized by considering the train operation mode of organizing multi-group and multi-routing. Taking peak hours as the research object, starting from the perspective of capacity and passenger flow matching, taking the best matching of line capacity and passenger flow as the optimization goal, comprehensively considering the constraints such as full load rate, maximum passing capacity, an optimization model of multi-group and multi-routing plan based on cross-sectional passenger flow differences is established. Taking Nanchang Metro Line 2 as a case study, the passenger flow and capacity analysis are carried out, and the model is used to optimize and solve it. Through comparison, it is found that the optimized operation plan significantly reduces the waste of transportation capacity and saves costs, thus verifying the good optimization effect of the plan obtained by the optimization model.

With the acceleration of urban construction and intelligence, the sources of multi-source data in cities have become more extensive, and the data from internal city management platforms has exploded. Multi-source data linkage is a method that enables the fusion of data from different sectors, which can help city managers to interpret comprehensive information about the objects of concern from multiple perspectives, and provide more scientific and accurate guidance for the management and development of the city. Based on the developed data management platform, through data pre-processing, multi-source data integration, matching and analysis, and the calculation of textual similarity of field descriptions of relevant subject data tables, it can effectively assist data management personnel to determine the correlation between data to support the subsequent multi-faceted data query and multi-dimensional information presentation.

Taking the road network in the third ring road of Xi 'an in 2020 as the research object, based on the space syntax model, the global integration degree and local integration degree are selected as the analysis indicators to analyze its characteristics. The results show that the global integration degree and local integration degree of Xi 'an roads are relatively high, showing the characteristics of high in the middle and low around the periphery, and the correlation between the global integration degree and local integration degree is highly correlated. At the same time, the selection degree is analyzed. Finally, the paper analyzes the spatial flow of the road network, and finds out that the traffic network of Xi 'an is basically reasonable and can meet the needs of travel. This paper hopes to provide some suggestions for future road traffic network planning in Xi 'an.

In order to improve the loading rate of the truck when the cigarette industry enterprises deliver goods and reduce the transportation cost. Starting from the idea of order carpooling, the problem solving involves the establishment of order carpooling model, cigarette loading, path planning, truck scheduling, loading execution and other optimization schemes. A Scientific loading system of multi-specification cigarettes truck under multi-constraint conditions was established to automatically distribute the order and generate the loading order with one click. Through system simulation, from 2019 to the first half of 2022, the average loading rate of trucks increased from 80.35% of manual distribution orders to 84.65% of system intelligent distribution orders, an increase of 4.29%. From the point of view of order grouping, this paper systematically combines the technical methods of cigarette cases loading and path planning, as well as truck scheduling and loading execution management methods, so as to improve the loading rate of cigarette transport trucks and reduce the transportation cost of enterprises.

In order to improve the management level of urban road traffic safety and promote the sustainable development of transportation, the systematic Decision Making Laboratory Analysis (DEMATEL) method is used to analyze the factors affecting urban road traffic safety. On the basis of sorting out the urban road traffic safety influencing factors in four aspects, such as people, vehicles, roads and environment, the DEMATEL model for systematic factor analysis is established; the direct influence matrix of systematic factors is determined and the comprehensive influence matrix of factors is calculated; the comprehensive influence matrix is used to calculate the degree of influence, the degree of influenced, the centrality and the degree of cause of each factor, to determine the causal attributes of each factor, and to find out the systematic. Finally, the effectiveness of the proposed DEMATEL model for system factor analysis is verified by taking an urban road transportation system as an example. The results show that eight key factors in the system can be identified using the model, including four key cause factors and four key effect factors. The constructed DEMATEL model is practicable, and the results of system critical factor identification using the model can provide a reference for decision-making related to urban traffic safety management.

The opinions and suggestions of diverse decision-making groups impact the formulation of regulations in the community. The grid/micro-grid system for community management fosters information exchange and enhances cohesion among residents. In this paper, the parameters of grid/micro-grid system are evaluated and integrated into a consensus model based on trust propagation and comprehensive clustering method. The model is employed to analyze the acceleration of consensus in a grid/micro-grid organized community. The experimental results illustrate that the grid/micro-grid system can effectively accelerate the consensus among decision-making groups.

With the increasing abundance of smart city construction cases, the essence and technical requirements of smart construction need to be continuously iterated and updated in conjunction with the latest information technology. Building Information Modeling (BIM) and Geographic Information System (GIS) technologies, widely applied in the fields of architecture and urban management, can provide more comprehensive data sources for smart city construction and compensate for the shortcomings in macro and micro data collection. This paper first explores the integration of BIM and GIS technologies, and then analyzes the application achievements of this integration from the perspective of smart city construction, using a specific international project as a case study. The findings provide reference for the construction of smart cities and the improvement of urban intelligent management and informatization.

With the rapid development of electric vehicles, the safety issues caused by overheating of power batteries are increasing, and even affecting the users' lives. Based on this, the state has issued relevant policies requiring measures to inform users before power batteries thermal runaway occur. In this paper, a set of early warning scheme for thermal runaway of power batteries based on DCF is designed. The core of this scheme is the monitoring algorithm of DCF integrated power batteries system, which combines the relevant information of DCF to actively push early warning information to vehicle users and restrict related activities within a certain period of time. In addition, a maintenance manual is generated for risky vehicles for maintenance personnel to use, so as to reduce the risk of thermal runaway caused by power batteries in the above way.

The development of big data has greatly improved the convenience of residents' lives. Among them, spatiotemporal big data has effectively improved the timeliness, practicality, and accuracy of geographic spatial information, and has been widely used in the field of public service facilities. Research has employed web crawling technology to obtain various city cultural facilities and residential community POI data, and has used self-compiled PYTHON code to obtain the travel time required for each community to travel to various cultural facility points by car. By combining data from the statistical yearbook, this study takes Xuhui District in Shanghai as an example and explores the adaptability of the supply and demand of urban cultural facilities from two dimensions: spatial adaptability and supply-demand relevance. Using methods such as nearest neighbor analysis, multi-distance spatial clustering, coordination location quotient, and correlation analysis, it was found that the spatial form of the distribution of cultural facilities and residents is similar, but there is a structural mismatch between supply and demand. The research results are helpful for optimizing facility configuration and promoting fine-grained urban management

According to data from the Maritime Administration, the majority of collisions at sea are caused by human error and failure to comply with the International Regulations for Preventing Collisions at Sea (COLREGs). To this end, an algorithm is designed to make collision avoidance decisions among multiple vessels according to COLREGs. On the premise of fully complying with the International Regulations for Collision Avoidance at Sea and seafarers' common practice, the algorithm constructs a decision-making model for collision avoidance actions during multi-ship encounters by using the true-motion graphing model and the collision avoidance method of the key ship according to the principle of analytical geometry. The results show that the algorithm can quickly determine the avoidance relationship between ships and solve the avoidance action amplitude and timing according to the requirements of the Rules and seafarers' common practice, which realizes the unity of the avoidance decision and action, and verifies that the algorithm proposed in this paper has the features of fast operation speed and good robustness.

With the continuous development of automotive active safety technology and the publication of the China New Car Assessment Programme (C-NCAP) Version 2021, active safety testing of longitudinal targets has received attention from vehicle manufacturers and testing institutes. In this paper, we propose an active safety test method for longitudinal targets based on a dual-band arrangement using an autonomous driving robot, a combined inertial guidance system, a digital pick-up device and a longitudinal target dual-band arrangement device, taking into account the actual road conditions. After practical tests, the feasibility of the method is verified, which has important implications for vehicle manufacturers and testing institutes.