This PDF file contains the front matter associated with SPIE Proceedings Volume 12940, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Hydraulic energy support is a structural part of hydraulic energy, and its fatigue life affects the service life and reliability of hydraulic energy. In this paper, the fatigue life of energy support under random vibration is studied by analyzing the condition of random vibration fracture. According to the fracture analysis results of the energy support, the fatigue life analysis model of the energy support is established, and the optimization and improvement direction of the energy support is proposed, which provides a theoretical basis for increasing the limit of the fatigue life of the hydraulic energy support under random vibration.

Traditional Kalman filter algorithms used for predicting drone positions rely solely on the historical positioning information of the drone itself, which makes it challenging to accurately estimate the position of the drone over longer time periods. In this paper, an improved IMM-KF (Interacting Multiple Models-Kalman Filter) algorithm is proposed. This algorithm incorporates a priori information about the drone’s planned flight route and the limited states of the drone, such as position, velocity, and acceleration, to achieve real-time prediction of drone position information. The advantages of interacting multiple models are also leveraged in the algorithm. Mathematical simulation results validate that this improved algorithm outperforms traditional Kalman filters in predicting drone positions, with the improvement becoming more pronounced as the prediction time horizon increases.

Automatic guided vehicles have widely utilized in the transportation works and assist to move the objects into certain areas. Therefore, avoiding the conflict of these vehicles and determine the effective scheduling strategies are an essential component to improve the effectiveness and guarantee the safe of these vehicles. Existing scheduling strategies are concentrated to utilize the mathematical models or train a machine learning algorithm to dispose the scheduling for each vehicles. In this work, we utilize the artificial neural network to solve the scheduling issue and achieve the high effectiveness of moving vehicles. Initially, we order the vehicles with the principle of priority and get the minimum Euclid distance for these vehicles. Subsequently, an artificial neural network is trained to obtain the scheduling routes for each vehicles. From our extensive simulation results and comparison analysis, we can conclude our proposed model can effectively achieve the scheduling routes for each vehicles and avoid the conflict or congestion among guided vehicles with reasonable computation and response time.

Oil-air lubrication ECT differential electrode sensor is established; the influence of isolated electrode and differential electrode on ECT sensor is analyzed; the range of The range of three structural parameters for measured electrode and differential electrode spacing, electrode length and electrode thickness is determined, and optimization is performed using Design-Expert. The results showed that the image relative error decreased by 18.4% and the image correlation coefficient increased by 36%.

Current synchronous control techniques for non-linear feedback in permanent magnet synchronous motors have large synchronous control time delays due to weak sensor signals. For this reason, a reinforcement learning based synchronous control technique for non-linear feedback of permanent magnet synchronous motors is proposed. A full speed position sensor is set up. Design of a master control centre based on reinforcement learning and reinforced processing of sensor signals. Design of non-linear feedback controller. Implementing synchronous control of permanent magnet synchronous motors based on feedback information. Simulation experiments have shown that this technology has a short response time and can significantly reduce the synchronization control delay of permanent magnet synchronous motors, with an average delay of 0.56 seconds. It has high practical application value.

In order to solve the problems of inaccurate position judgement and insufficient sense of presence in the process of remote operation of mobile robotic arms via video in the tunnel environment, a positioning system based on ultra-wideband positioning technology and a sensing system based on the combination of 2D LIDAR and pitch rotary table are proposed, and a verification platform is built and experimentally studied. The research in the thesis proposes a new solution to the real-life problems faced by engineering technology and has certain research value.

Aiming at the requirement of low cost, high precision and high stability of space advanced technology demonstration satellite, a method of heterogeneous backup of attitude control sensor is proposed, and the design principle of the sensor is analyzed in depth multi-sensor information fusion algorithm is designed. The control strategy of the satellite is optimized, the weighted PID control strategy is designed, and a reasonable integral separation threshold is selected. Control bandwidth is adjusted to reduce the control deviation caused by time delay. The objective of the attitude control system is to achieve control accuracy better than 0.1°(3σ) and control stability better than 0.003°/s(3σ).Through the analysis of on-orbit test data, the control accuracy of the satellite is better than 0.005°(3σ), and the control stability is better than 0.0005°/s(3σ).

In this paper, we present a novel method for implementing three-loop control of motors using digital signal processors (DSP). We introduce a fast current loop algorithm (FCL) to enhance the control performance of the servo motors. The three separate loops include the position, speed, and current control respectively. We implemented the three-loop control of motors using a TI C2000 series DSP and conducted laboratory testing. By comparison with traditional control methods, the proposed control method exhibits higher response speed, lower overshoot and greater resistance to external disturbances and system parameter changes. Hence, this method can be widely adopted in high-performance servo motor control applications

To improve the efficiency and safety of wood handling, an AGV magnetically guided handling unmanned vehicle is designed. The AGV magnetic navigation unmanned vehicle uses the STM32 as the control system, acquisition of magnetic signals using the MGS-16FP magnetic navigation sensor, HC-SR04 ultrasonic distance measurement module collects distance information, MFRC522 module reads and writes line information, LCD shows unmanned vehicle status. The ESP8266 module transmits data via the MQTT protocol for path planning and real-time monitoring. The AGV magnetic navigation unmanned vehicle designed in this paper has high efficiency and stability and can work stably indoors to automate wood handling, which greatly improves the efficiency and safety of wood handling.

The trend of China's automobile policy control is changing to ex-post supervision, and the emission control of in-use vehicles is gradually emphasised. In order to address the common SCR (Selective Catalytic Reduction) cheating technology in the market, this paper uses cheating and normal OBD (On-Board Diagnosis) data collection with different engine models as data sources, and develops three kinds of vehicle SCR cheating models based on OBD data using BP neural network, logistic regression and support vector machine, respectively, and verifies the accuracy. It was found that the support vector machine model is suitable as a test model due to its high accuracy, high generalisation performance, and simple and fast computation.

To improve the penetration capability for hypersonic vehicles, a rapid trajectory planning method for hypersonic penetration based on sequential convex optimization and deep learning is proposed in this paper. Firstly, to get a better evaluation of the hypersonic penetration problem under two interceptors, the penetration dynamics and its convex model are presented, which include the linearization, the discretization and the convexity processing. Meanwhile, to realize the rapid hypersonic penetration trajectory planning, the penetration problem is transformed into a second-order cone programming problem. Secondly, in order to further improve the real-time performance of hypersonic penetration trajectory planning, a penetration trajectory generation method based on deep neural networks is proposed. Finally, multiple scenarios are simulated to verify the proposed DNN based rapid trajectory planning method. Results indicate that the method can realized the effectively penetration which both realize the milliseconds calculating and the excellent robustness.

The construction of rural distribution network is limited by investment, resources and policies and has not been effectively developed, which leads to the weak functional capacity of distribution network and poor power quality of distribution network. Based on this, this paper first analyses the important value and function of power quality monitoring, then studies the design of rural distribution network monitoring system on account of power quality governance, and finally gives the utilization effect of rural distribution network monitoring system

In low-voltage AC-DC hybrid microgrid, power fluctuation is caused by load switching. In order to ensure the stable operation of microgrid, a control strategy for interconnecting converters in low-voltage microgrid is proposed. Aiming at the low-voltage microgrid, the difference between its operation characteristics and the traditional microgrid is analyzed, and the topology structure and sagging characteristics of the low-voltage microgrid are studied. The characteristics of AC and DC side voltages are obtained by standard processing on both sides of the subnet voltage, and the working interval of the interconnection converter is redivided to cope with the power transmission direction under different subnet states. For the frequent switching problem that may occur in the working interval of the interconnected converter, the action threshold is added to control its switching state. Finally, a simulation model is established in Matlab/Simulink to verify the feasibility of the proposed control strategy.

Aiming at the path tracking problem of unmanned surface vehicle, a sliding mode control algorithm enhanced with adaptive control technology is proposed. Adaptive control technology is used to predict the boundary value of unknown external environment interference. In order to improve the robustness of the system and the convergence rate of path tracking, the optimization algorithm is designed based on the hyperbolic tangent function control reaching law. A new double power combination function is proposed as the control law to design the sliding mode control algorithm to control the forward and steering of unmanned surface vehicle. The chattering phenomenon caused by sliding mode control is solved effectively and the convergence speed of path tracking is improved. The curve fitting method is utilized to realize the effective tracking of any planned flight path. The simulation results demonstrate that the proposed adaptive sliding mode control algorithm is able to ensure unmanned surface vehicle track the upper reference path accurately under the time-varying disturbance of wind, waves and currents. In additional, the adaptive sliding mode control algorithm embedded an approach control law, called new double power function, is confirmed that has strong robustness and fast convergence rate of path tracking.

Taking the downshift process of one dual clutch transmission as the research object, based on the correlation analysis of subjective and objective evaluation, one evaluation standard of shift noise with the vibration acceleration of mounting point as the evaluation parameter was proposed. At the same time, the types and mechanism of gear shift impact noise are analyzed, and the strategies to improve gear shift impact noise are proposed.

Pneumatic flexible joints are widely used in agriculture, medical and environmental detection and other fields due to their advantages of flexibility and human-computer interaction safety. At present, the stiffness of the pneumatic flexible joint is insufficient and the driving ability is weak. Based on the principle of particle interference, a variable stiffness elastic shaft is proposed. The mathematical model of the variable stiffness elastic shaft is established, and the relationship between the stiffness and air pressure of the elastic shaft is analyzed, and the relevant experimental verification is carried out. The experimental results show that the elastic shaft has simple structure and good variable stiffness , and can be used as a variable stiffness part in bionic machinery such as agriculture, medical rehabilitation and service.

The analysis and evaluation of GNSS performance under ionospheric scintillation has always been a critical concern in the modern aviation industry. This paper based on the ionospheric scintillation model and satellite navigation solution principles, the GNSS aviation service performance assessment software is designed and applied to different ionospheric scintillation intensities, and the C/C++ language is used as the programming language to realize the GNSS aviation service performance assessment functions under different ionospheric scintillation environments and different flight phases. The software can flexibly change the ionospheric scintillation intensity, flight trajectory can be controlled. The software is highly scalable and reusable, and it is compatible with GPS/BDS/GALILEO/GLONASS systems. It fully considers the specification and visualization of the GNSS aviation service performance assessment under ionospheric scintillation conditions, promoting the improvement of the GNSS aviation service performance assessment system.

Aiming at the problem of prime mover speed variation caused by frequent load variation of Marine diesel generator set under actual operating conditions, a speed control system of Marine diesel generator sets based on DSP is designed, the speed control principle of Marine diesel generator sets is analyzed, and the hardware block diagram and software flow of the system are designed. In terms of hardware, the TMS320F28335 chip is used as the core components. Regarding the software aspect, a model for the marine diesel generator set has been simply constructed. Serial port to Ethernet interface is used to complete the communication between DSP hardware environment and Simulink environment, and the hardware in loop semi-physical simulation of the system is realized. By adjusting PID parameters, the system simulation waveform changes were analyzed. Finally, the DSP controller performance is verified by increasing and reducing the load abruptly. The results obtained from the simulation indicate that the speed control system possesses high response speed, good stability and strong tracking ability

Aiming at the strong coupling of speed and heading control of unmanned sailboats, this study proposes joint control of the speed and heading of unmanned sailboats based on nonlinear model predictive control (NMPC). Different from the existing control methods, which only carry out maximum speed and heading separation control, this study proposes the joint control of unmanned sailboats at any speed based on NMPC. A 4-DOF dynamic model of the unmanned sailboat is used, and the roll angle is limited to prevent the unmanned sailboat from capsizing. Simulation experiments prove that the proposed method can realize the speed and heading joint control of unmanned sailboats, and the roll angle can be limited, preventing the unmanned sailboat from capsizing.

Based on the Proximal Policy Optimization with clipped objective (PPO-clip) algorithm framework, an autonomous maneuver decision-making method for short-range 1v1 unmanned combat aircraft vehicles (UCAVs) is designed and implemented. In this paper, the curriculum learning (CL) mechanism is used to train the maneuver decision-making model to solve the problem that the model cannot converge when fighting against complex maneuvering enemy UCAV. The entire training process is divided into 4 stages to fight against enemy UCAV, which maneuvers range from simple to complex, and finally achieve our UCAV against the enemy UCAV with intelligent maneuvers. Through four groups of simulation experiments, this paper proves the effectiveness of the PPO-clip algorithm and the curriculum learning mechanism that can speed up the convergence of the model.

A fault diagnosis method for rolling bearing based on improved Whale Optimization Algorithm (WOA) is proposed to address the nonlinearity, non-stationarity, and unclear fault characteristics of the rolling bearing vibration signals. Firstly, adaptive inertia weight, nonlinear convergence factor, and Levy flight theory are introduced into WOA, and an improved WOA algorithm based on a hybrid strategy is proposed to enhance the optimization speed and accuracy. Then, fuzzy entropy is used as the fitness function of the WOA algorithm to optimize the parameters of Variational Mode Decomposition (VMD), obtaining the optimal combination of the number of modal components and penalty factor. Finally, the optimized parameter combination is applied to perform VMD on the bearing fault signals, yielding several modal components. The improved WOA algorithm effectively avoids mode mixing in the optimized VMD decomposition, whereas Empirical Mode Decomposition (EMD) exhibits mode mixing. This further validates the superiority of the proposed method. The research results provide an effective improvement approach for existing rolling bearing fault diagnosis techniques.

Air combat target tactical intent refers to the analysis and inference of the enemy's combat intentions in real-time, adversarial environments by extracting battlefield environmental information, static attributes, and real-time dynamic information of air combat targets, combined with knowledge from the military domain. To achieve this goal, many machine learning-based methods have been proposed to infer aircraft intentions. However, these methods are only applicable to individual aircraft and cannot predict the intentions of the entire formation. Therefore, we propose an attention-based multi-level LSTM model that incorporates multiple levels and attention mechanisms to enhance the focus on key information and improve prediction efficiency, resulting in promising experimental results.

Accurate prediction of vehicle lane-change behaviour is beneficial for traffic efficiency and safety. This paper incorporates driver behavioural information indicators into the prediction of lane change behaviour. A driving simulator was used to collect driver behavioural characteristics and vehicle operating parameters during lane keeping and lane changing, and through theoretical analysis and data testing, seven indicators were obtained that could be used to predict lane changing behaviour: target area gaze time, head level turning angle, vehicle speed, lane drift angle, distance to the vehicle in front, and time to collision. Using the experimentally collected variable data, a logistic prediction model was developed and validated to have satisfactory prediction accuracy.

Model Predictive Control (MPC) combined with the planning layer can realize the autonomous obstacle avoidance of vehicles. Still, the generated trajectories are often not smooth enough and have considerable tracking errors, resulting in poor stability and comfort of cars. To address this problem, a PID-MPC dual closed-loop intelligent vehicle controller that can reduce tracking error and improve vehicle stability and comfort is designed. First, the differential term of the traditional PID control algorithm is introduced into the first-order inertial filtering link to enhance vehicle stability and reduce tracking error. Secondly, based on the nonlinear model of vehicle dynamics, the MPC controller was used as the outer control loop, the improved PID controller was used as the internal control loop, and the dual-loop controller solved the optimal control sequence to realize the dual closed-loop optimal control of the intelligent vehicle. Finally, the feasibility and effectiveness of the designed PID-MPC dual closed-loop controller were verified by joint CarSim/Simulink simulations. The simulation results showed that the tracking trajectory generated by the PID-MPC controller was smoother, the front wheel steering angle, yaw angle, and lateral acceleration changed more smoothly with smaller amplitudes, and the car's comfort and stability both increased.

As the core of autonomous vehicle technology, perception chips are essential for data processing and environmental sensing. The efficacy and reliability of these chips are strongly impacted by the various functional risk factors may confront in scenarios in the real world. This paper provide efficient preventive measures and an innovative method for addressing functional safety failures in perception chips across a variety of real-world scenarios. It systematically investigates typical failure scenarios in a variety of scenarios as well as offers a detailed explanation of the underlying difficulties. A range of novel functional safety testing techniques are introduced in the study, which builds on this basis and allows for quicker and more accurate fault processing. The study additionally explores recent developments and preventative methods to enhance the overall safety functionality of autonomous systems. The study conducts extensive simulation experiments and comparative analyses to show off the viability of the proposed approach, showcasing the notable improvements in terms of fault recognition rate and processing speed. Results from experiments support the method's significant advantages for enhancing the functionality in perception chips, leading to improved functional safety throughout the autonomous driving system.

In recent years, intelligent vehicles have been at the center of research in automotive engineering and the future direction of the automotive industry, and are gradually showing a trend towards practicality. Excellent environmental perception and detection capability is the prerequisite and foundation for the normal operation of driverless vehicles, which directly affects the realisation of subsequent planning and decision-making, control execution and other advanced functions. This paper introduces the development status of intelligent vehicles at home and abroad, studies the current research status of mainstream perception algorithms, analyses the technical challenges encountered in perception for intelligent vehicles in urban roads, and illustrates the development goal of establishing efficient and reliable environmental perception algorithms in complex scenarios.

In this paper, we propose a dynamic model of an asymmetric passive walker with the knee joint. The stable initial conditions of the dynamic model are obtained by the cell mapping method. We analyze how the length ratio of the left and right legs of the walker (the left and right leg mass and the moment of inertia change along with the length of the leg) and the angle of the ramp influence the gait characteristics of the walker, such as the length of step, gait cycle, and walking speed of the model. Based on the existing researches, if there is a very small asymmetry difference, the walker will have a double-step gait. When the leg length difference reaches a certain level, a four-level gait will appear, and as the leg length ratio continues to change, eight times and sixteen times will continue to be shown out, which in turn evolves into a chaotic gait.. If the proportion of leg length exceeds an upper bound, a stable passive walking gait will no longer be found. According to the study on the asymmetric passive walker, this research can not only guide the design and manufacture of a driven biped robot but also be used for clinical-pathological asymmetric gait treatment, exoskeleton assist device development, prosthetic design, and prosthetic adaptability training to provide theoretical guidance.

The automobile transmission box has a complex structure and many features. To improve the design efficiency of automobile transmission, shorten the production cycle and meet the personalized requirements of products, the rapid variation of automobile transmission was studied, and a method of rapid variation of automobile transmission based on feature modularization was proposed. This method introduces the idea of feature modularization, modularizes the model features, then builds the transmission variation template based on the construction of the module sketch, combines it with the relevant software modeling methods, and finally builds the transmission rapid variation platform with the NX secondary development technology. The effectiveness and reliability of this method are verified by a concrete example.

In order to obtain the inherent law of the structure and performance of the SMA actuator of the bionic flexible dexterous hand. Four phase transition temperature parameters of SMA wire were obtained by temperature test. The Brinson constitutive model was used to describe the material relationship of SMA. The finite element model of SMA mechanical properties was established by user material subroutine (UMAT). The validity of the secondary development was verified by comparing the results with the literature. Based on this, the parametric finite element modeling and simulation of SMA actuator were carried out, and the influencing factors of mechanical properties of SMA actuator were analyzed. The results show that the SMA wire diameter of the dexterous hand actuator has a great influence on the performance of the actuator, while the structural parameters such as the winding spacing and the actuator diameter have little effect on the driving performance of the SMA wire. The pre-stretching force and output force of the actuator increase with the increase of the diameter of the SMA wire. The research results provide a reference for the design of related SMA actuators.

The anti-slip strategy of the P2.5 hybrid system is based on the anti-slip condition determination, the anti-slip torque limit control and anti-slip torque filter control and the clutch coordination. The specific contents are as below: The anti-slip condition determination. The maximum allowable requested torque limit is calculated in advance according to the current ambient temperature, slope, steering angle, speed and other conditions for anti-slip feedforward control, so as to avoid the vehicle slip on the low adhesion road surface. The anti-slip torque limit control. When the body stability system works normally, the anti-slip control is handed over to the body stability system for control. When the body stability system is not working properly, the vehicle controller performs closed-loop control according to the slip error. According to the speed difference between the front wheel and the rear wheel, the acceleration of the left front wheel and the right front wheel, the anti-slip grades are judged and the anti-slip torque is limited based on the difference between the target speed of the output shaft and the actual speed of the actual output shaft and the change rate. The anti-slip torque filter control. In order to rapidly reduce the wheel end driving force to suppress the slip, the torque filtering speed is fast enough to meet the requirements, and the torque filtering at this time needs to be distinguished from the torque filtering at normal times. In order to avoid the frequent jump of the torque filter coefficient caused by anti-slip, the torque filter coefficient calculated by anti-slip is used only when the requested torque and anti-slip torque limit are greater than the threshold. The vehicle results are conducted and show that the anti-slip control strategy can meet the request of the driver and the good drivability performance can be achieved.

A new design of the suspension system, the third spring decoupling suspension, is made to improve the handling stability and ride comfort of formula student racing. According to the requirements of race rules [1], the suspension performance parameters were calculated combined with the basic parameters of related racing cars. Based on CATIA software, three-dimensional design and assembly of suspension components were carried out. And the dynamics simulation analysis and optimization of the suspension system were carried out to further improve the performance of the racing car based on ADAMS software.

This paper provides a better living and learning environment for training parrots. During the training period, the parrots were trained on a daily basis, provided with food when needed, and regularly disinfected and cleaned up their habitat. Reduce the workload of trainers and improve the efficiency of parrot training. Through the single-chip microcomputer controller, the use of temperature detectors and humidity detectors to monitor the environment, and according to the actual environmental conditions to adjust the temperature and humidity, displays the temperature and humidity of the current environment and time on the OLED. To arrive at feeding time, control the motor for feeding, or use an LED lamp to simulate the expression. Arrive disinfect clean time, control motor to undertake disinfect and clean excrement, or use an LED lamp to simulate to express. To reach the preset time, play voice to teach the parrot. The app monitors the environment on the phone and can directly control feeding, garbage removal and voice teaching functions. Finally, the correctness and reliability of the design are verified by simulation.

Since the trash can is not sanitary and fails to do the garbage classification, a household intelligent trash can based on SCM is designed. Firstly, the hardware of the intelligent trash can is designed, whose function includes garbage classification, voice tracking, UV disinfection, and other hardware circuit design. In this design, a speech recognition module is used to classify trash cans, UV disinfection lamp components are employed to achieve the disinfection and sterilization of garbage viruses, and the tracking function is used to realize the free movement of trash cans, providing great convenience for the intelligent trash cans. Secondly, combined with software design, the system functions are further improved to make the trash can more intelligent and practical.

As modernization advances, people demand a better life with the prime demand for building a living environment, which is simple, convenient, intelligent, environmentally friendly, safe, and reliable. In this paper, a smart home temperature control system is designed, with the Internet of Things technology as the core. It achieves control through information collection, data upload, command delivery, and automatic control. The network protocol is used to upload data information to the cloud platform in real-time, so as to realize the overall intelligent home temperature control system design, which can solve the problems of uncomfortable indoor temperature and humidity, high carbon dioxide concentration, high methane concentration, etc. It provides supplementary and technical support to realize intelligent urban life and adapt to modern, fast-paced urban life.

The throwable robot can shift from a spherical robot to a mobile robot and perform the environmental survey mission while entering the working environment in this manner. In order to enhance the robot's environmental adaptability, a motion control system is proposed to address the issue that the driving wheel excessively slips while the robot drives on the soft ground. The control system consists of kinematic controller, dynamic controller and slip rate controller. Kinematic controller realizes track tracking. The mechanical model of wheel-terrain interaction is introduced into the design of dynamic controller to obtain the target slip rate while tracking the target speed. To avoid the drive wheel slipping excessively, the target slip rate is limited to a certain range. The slip rate controller can track the target slip rate. The simulation results show that the robot can successfully track the straight track and circular track and control the slip rate of the driving wheel in a limited range, which proves the effectiveness of the control method.

With the rapid progress of science and technology, an increasing number of robots are used in society. Compared with multi-legged crawling robots, humanoid robots have gained more flexibility and reliability with a structure similar to the human body, and have been widely used. Therefore, the proper pose control of the humanoid robot is particularly critical. This paper proposes and designs a humanoid robot control system based on human pose detection. The system detects the posture image of the human body by using the OpenPose method, and then uses the space vector method to calculate the rotation angle information of the human body joints. After being processed, the information is sent to the steering gear controller by the Raspberry Pi. After the controller solves and processes the signal, the pose of the robot is controlled by driving the steering gear

This paper aims at the supply chain inventory management system with information sharing delay, and the design problem of distributed model predictive controller under communication fault is studied. Firstly, the inventory model of information sharing delay is constructed as a state space model under communication failure. Based on model predictive control and state feedback control method, a distributed model prediction controller is designed in this paper. Secondly, this paper uses the controller to optimize the change of the quantity of ordered goods, reduce the management risk of inventory, and achieve the goal of adjusting the inventory level to a safe position. Finally, a simulation living example is given to verify the availability of the proposed control strategy.

The traditional fast rapidly exploring random tree(RRT) algorithm has the advantages of probabilistic integrity in the motion planning of robotic arm, but at the same time, it suffers from blind search, poor goal orientation, and slow planning speed, while the goal-oriented RRT algorithm based on probability P has good goal orientation but suffers from the problem that it is easy to fall into local minima that cannot be jumped out. In this paper, we propose an RRT algorithm with variable probability P based on the influence of historical nodes to address the above problems so that it has the advantages of both algorithms, relying on the ability of historical nodes to perceive the environment, allowing it to maintain a high goaldirected probability in an open environment and being able to automatically reduce the value of P in an environment with many obstacles, giving it the ability to jump out of the local minimum of traditional RRT.

Underwater vehicles have highly non-linear, strongly coupled, multi-degree-of-freedom and time-varying dynamics models, and work in complex environments that require a high degree of adaptivity. Its motion control has always been a very complex challenge. In order to solve the complex communication methods of the traditional Modbus TCP, The three handshakes during TCP protocol data transfer incur unnecessary time expenses, while solving the core problems such as improving controller performance. Based on the original Modbus TCP, the Modbus UDP communication method is proposed to provide an efficient and readable communication mechanism for underwater vehicle communication, which is more effective in sending commands and obtaining feedback to the main control board of the lower computer. It also uses QT to design a display interface that is easy to listen to downstream data and display feedback data. The proposed Modbus UDP communication method is experimentally verified in an AUV (Autonomous Underwater Vehicle), and through the verification, it is concluded that the communication method is reliable, readable, easy to implement and highly portable, demonstrating the applicability and effectiveness of the improved Modbus UDP communication method for AUV control.

To solve the problem of uneven flow distribution of constant humidity gas in the gas pipeline network, this paper uses FLUENT to simulate the flow field and pressure of the butterfly valve and concludes that the butterfly valve can be used to regulate the flow distribution. Based on the above conclusions, the authors designed a butterfly valve controller, which can drive the butterfly valve to open and close through the driver chip to regulate the flow in the pipeline network. Finally, through experimental verification, it is shown that the butterfly valve controller can adjust the flow resistance coefficient of each branch by controlling the opening and closing degree of the butterfly valve, and the most suitable adjustment range of the opening and closing degree of the butterfly valve is 20°-40°.

For the adjustment and design of weight factors in the model predictive torque control (MPTC) of brushless DC motor (BLDCM) is too tedious and full of uncertainties. In this paper, a modified whale optimization algorithm (IWOA) is utilized to de-regulate the parameter weight factors in real time. The algorithm incorporates chaotic mapping and greedy selection strategies, where chaotic mapping enables the algorithm to achieve population diversity in the initial stage, and greedy selection strategy enables the algorithm to update the local in the case of falling into a local optimum, making the algorithm more capable of finding the best and more accurate in optimizing MPTC. In order to verify whether the proposed approach can reduce the speed error and torque fluctuation, by modeling and conducting simulations to verify the torque and speed of the motor, and taking into account the switching frequency loss factor, the results show that the torque pulsation is further reduced, and the speed error is further reduced in the proposed approach compared to other intelligent algorithms, confirming the effectiveness of the proposed approach.

Smart homes are rapidly expanding, enabled by new technologies such as the Internet of Things and artificial intelligence. The scope of smart home technology continues to grow and encompass more areas. Nevertheless, the expansion of smart home like a wider population is hampered by the fact that manufacturers are unable to keep up with the development of intelligent technologies and the high cost of the corresponding research and development. This paper proposes to use digital twin technology for smart home system. In order to promote user experience and reduce the cost for smart home, it also puts forward to use BP network technology to build a smart home system model to ensure the integrity and stability of the system, and illustrates the emergence of more mature smart lighting systems today. This paper elaborates in detail how to implement the intelligent lighting system and the advantages that the proposed lighting system can bring to people's lives.

With the rapid development of artificial intelligence, big data, image recognition, virtual reality technology, voice control technology and other emerging technologies, There has had a certain technical basis in human-computer integration. Human-robot integration is the essential feature of next-generation robots, and the development of human-robot integration technology is of great significance to the future of China's robotics industry. This paper reviews the current research status of related technologies in human-machine co-integration-related fields, analyzes the focus of mobile robot control research in human-machine co-integration environments, and outlines the aspects of path planning and control methods involved in the control of multiple mobile robots in human-machine co-integration environments. Finally, the prospect of cooperative control of multiple mobile robots in human-robot co-integration environment is presented.

An improved sliding mode control (ISMC) method based on linear extended state observer (LESO) is proposed to solve the trajectory tracking problem of underactuated ships with unknown states in the environment of disturbance. It can limit the input in a given range and has strong anti-saturation ability. Firstly, the ship kinematics and dynamics models are transformed into integral series models by coordinate transformation, and then the unknown states are estimated by LESO. Secondly, the corresponding virtual control law is designed by backstepping method. Then, the control frequency of the ship's main propeller and rudder device is reduced by ISMC, and the problem of differential explosion caused by high order derivative of the state quantity is avoided by passing the state quantity through the first order filter. Aiming at the actual input limitation problem of the controller, the anti-saturation strategy is introduced. The effectiveness of the control law is proved by Lyapunov theory.

In order to solve the internal coupling problem of the levitation module, it is necessary to design decoupling algorithms for the levitation system under the modular structure to suppress the coupling disturbance between the levitation points of the levitation module. In this paper, we take the levitation module as the object of study and carry out the decoupling study of the modular control of the levitation system. We design the linear active disturbance rejection control and the feedback linearization as the decoupling controller of the system respectively, and analyse the dynamic performance of the levitation system in the two kinds of decoupling control respectively. The simulation results show that the decoupling controllers designed by active disturbance rejection control and feedback linearization methods respectively can solve the internal state coupling problem of the levitation module and improve the levitation control performance.

In order to improve the stability of the hydraulic control system for coal mining equipment, PWM technology is used to accurately control the hydraulic proportional valve group. A communication system is established through CAN communication, and PWM signals are used to directly drive and control the proportional solenoid valve. The control accuracy of the electro-hydraulic proportional valve is ensured by feedback on the output current of the controller. The characteristic curve and time response of the valve are parameterized and programmed to improve system performance, Improved system quality and load application scenarios. The actual operation results show that the performance of the mining equipment walking electro-hydraulic control system is stable and reliable, which improves the safety, reliability, and sensitivity of the equipment walking electro-hydraulic control. This lays a solid foundation for achieving unmanned operation of mining equipment and building green mines.

A method integrating rapidly-exploring random tree (RRT) and improved dynamic window approach (DWA) is proposed to realize obstacle avoidance of mobile robots in complex dynamic obstacle environments. Based on the known environment information, the global optimal safe path is generated using the improved RRT algorithm. To address the problem of too many invalid nodes in the expansion of the RRT algorithm, the efficiency of the algorithm is improved by introducing a bidirectional alternating search (BAS) strategy; the forward search tree and the backward search tree alternate searching paths with new nodes as the target nodes until the paths meet; and the redundant nodes in the paths are removed using path node filtering to shorten the path. Secondly, the DWA algorithm is used to track the improved RRT algorithm to plan the path. When unknown obstacles appear in the environment, the weights of the DWA algorithm evaluation function are dynamically adjusted by fuzzy control to avoid the obstacles and return to the original route in time. Finally, the simulation experiments verify that the proposed improved fusion algorithm in complex dynamic environment has short running time, small path cost, and always keeps a safe distance from the obstacles to ensure the optimal path is tracked while safely avoiding the dynamic obstacles.

Traditional warehouse logistics has faced huge challenges with the development of China’s information technology industry. Intelligent logistics, especially intelligent warehousing, has become a hot research topic worldwide. Amazon Kiva system, a representative of intelligent warehouse handling systems based on multiple mobile robots, has revolutionized the warehousing industry with its “goods-to-man” picking model, which reduces the time and labor costs of finding goods. However, this model still has many limitations, such as low efficiency and scalability, when dealing with large-scale picking tasks. In this paper, we propose a novel intelligent warehouse handling system that overcomes these limitations and improves the system performance. We present the system architecture, design principles, and key algorithms of our system, and evaluate its effectiveness and robustness through simulations and experiments. Specifically, we propose an order task scheduling algorithm based on order adaptation and discrete particle swarm optimization, and a path planning algorithm based on 2D planar graph for the warehouse scenario. We show that our integrated algorithm can improve the system efficiency by solving the two key problems of order scheduling and route planning in the smart storage system.

With the rapid development of mobile robot technology, bipedal wheeled-leg robots, which combine the efficient mobility of wheeled robots and the adaptability of legged robots to uneven terrain, are gradually attracting widespread attention. In complex terrains and unstructured environments, such as stairs, ramps, and narrow spaces, bipedal wheeled-leg robots have superior mobility and adaptability. This article introduces a motion control method for bipedal wheeled-leg robots with a five-bar leg configuration. This method decouples the robot system based on a distributed model, analyzes the body and wheeled leg system separately, and implements a controller for the robot's rapid movement on flat terrain, adaptation to unstructured terrain, and jumping movements through virtual model control and model predictive control. A series of experiments have verified the feasibility of the prototype system's movement in real-world scenarios.

This paper studies the globally optimal coordinated control problem of distributed autonomous underwater vehicle (AUV) formation. Remarkably, the global cost function and communication topology are pregiven, which makes the traditional Riccati-based strategy inapplicable. As the main contribution of this paper, an effective design strategy for optimal distributed controller is presented. Detailedly, the formation coordination control problem of multi-AUV system is first transformed into the consensus problem of AUV formation, and the consensus state of leader-follower AUV formation is defined. Secondly, considering the relative drift and offset of AUV clock and the bounded time-varying communication delays in the interaction network, a prediction-based formation consensus controller is proposed by introducing the clock synchronization strategy. Moreover, the stability of AUV formation with this controller is proved. Then, by deeply exploring the internal relationship between global cost and relative state errors, the fully explicit formula of the cost function with respect to gain parameters is derived. Based on this, the existence of the optimal distributed controller is further proved. Simulation results show that the proposed optimal controller is effective.

In order to better balance the driving safety and ride comfort of the car, a new two-stage semi-active suspension structure is designed with three elements of variable inerter, spring and damper as the suspension structure. Firstly, the ISD semi-active suspension is set up in Matlab/Simulink; Besides both the road model and fuzzy controller model are set up to simulate the dynamics of three schemes, namely passive ISD suspension, inerter PID control and inerter fuzzy control respectively. The results indicate that the performance of ISD suspension adopting the fuzzy control strategy is better, compared with the passive ISD suspension and the ISD PID-controlled suspension. thus it verifies the superiority of using fuzzy control strategy.

The control accuracy of Single Gimbal Control Moment Gyroscope (SGCMG) is limited by the existence of interference factors such as dynamic imbalances. This paper analyzes the output torque disturbance of SGCMG from a dynamic perspective and verifies the influence weights of disturbance factors through simulations, providing a reference for the control of torque output stability in SGCMG. Firstly, considering factors such as static and dynamic imbalance, installation errors of the high-speed rotor system using the Newton-Euler method, a preliminary analysis of each subsystem is conducted. Secondly, considering disturbance factors such as frame speed, a dynamic model analysis is performed. Finally, based on actual parameter values, the characteristics of the system's output torque are analyzed through simulations. Experimental results show that the torque output process of the system is stable under ideal conditions. However, in the presence of disturbances such as dynamic imbalance, the stability of SGCMG's torque output process is influenced to varying degrees. The analysis of the influence weights of various disturbance factors is accomplished using principal component analysis.

The quadruped robots have a good prospect in substation inspection applications because of their stronger terrain adaptability. This paper focuses on the gait planning and motion control of small intelligent quadruped robots. Firstly, the mechanical structure design of a quadruped robot is completed based on SolidWorks software. Secondly, the kinematics of the designed quadruped robot is analyzed, and gait planning of the walk gait and the steering gait is designed. Finally, the experiments of the walk gait and the steering gait are completed on the physical platform of the quadruped robot designed in this paper. When the robot performs the walk gait, the average forward speed is 0.04 m/s. When the robot performs steering gait, the average steering speed is 9°/s. The experimental results show that the gait control method designed in this paper can achieve stable control of quadruped robot motion.

Adaptive backstepping sliding mode control method is used to synchronize two Volta’s chaotic systems with external disturbances. The hyperbolic tangent function is adopted to replace the sign function in the sliding mode controller, to weaken the chatter problem caused by the sign function. Using Lyapunov theory and Barbalat's Lemma to prove the asymptotic convergence of the proposed control system. The numerical simulation demonstrates the effectiveness of the designed controller.

This paper presents the development of a model-based event-triggered control strategy for addressing the stabilization problem in nonholonomic robots. The proposed strategy is implemented in a practical setup. To handle the inherent nonlinearity of nonholonomic robots, feedback control is employed. Additionally, an event-triggered tactic is incorporated to alleviate communication burdens and computational loads, while ensuring system performance. The stability of the closed-loop system is rigorously guaranteed. Furthermore, comprehensive analyses of experimental results are provided, encompassing simulations as well as real-world experiments.

In this paper, the typical pressure changes inside vacuum tank is, through a pressure control system, abstracted by means of a simulation of an airtight test device for a vacuum pipe train; Due to the characteristics of the pressure control system–repeated simulations of typical pressure changes in vacuum tank, and the difficulty in building accurate mathematical models, numerical models of the mass-pressure conversion of the system have been constructed; the pressure control of the system is also studied by PSOILC (the particle swarm optimization iterative learning control) algorithm in accordance with the system characteristics. Simulation results show that the PSOILC performs well in terms of convergence rate and control accuracy. In summary, this algorithm can effectively improve the control accuracy, convergence rate, and dynamic performance of the control system.

In order to understand the design of substation inspection robot control system, a research on the design of substation inspection robot control system based on wireless video monitoring is put forward. Firstly, this paper introduces the overall flow of this embedded system software; The implementation process of PMAC motion control card driver is given. Secondly, the functional requirements of substation inspection robot are analyzed. Combined with the actual technical indicators, the overall architecture of the robot is designed for practical application scenarios, and a design scheme of inspection robot control system based on hierarchical design is given. Then, the core devices involved in the scheme are calculated and selected. Finally, it shows that the control system can really form a full monitoring mode, which can greatly improve the safety, reliability and real-time performance of substation equipment operation.

Traditional lung puncture surgery requires a very high level of experience and operational proficiency from the surgeon, and a slight inadvertence during the procedure can cause irreversible damage to the patient's lungs. At the same time, long-term exposure to X-ray radiation in the surgical CT room can cause significant radiation dose overload injuries to the surgeon. To address the above problems, this paper designs a novel robotic system for lung puncture surgery and investigates it for master-slave control. First, a mechanism that can deliver multiple medical devices simultaneously is proposed to enable rotation, delivery, clamping, and bending control of the medical devices; second, a method to manipulate the surgical robot using a touch panel is proposed; and finally, a combination of convolutional neural network and image processing is used to assist in controlling the automatic rotation of the endoscope to reduce the complexity of the surgery. The final animal experiment of the paper verifies the feasibility of the robot.

The nonlinear dynamic characteristics of a quadratic polynomial discrete chaotic system are studied. The system has three system parameters and presents different period-doubling bifurcation characteristics. In this paper, the bifurcation process of three different parameters as bifurcation parameters is analyzed. Meanwhile, the influence of the initial value of the system on the period-doubling bifurcation of the system is studied. The translation method is used to control the bifurcation of the system. Changing the translation control parameters can change the dynamic characteristics of the system. By taking different values of the translation control parameters, the bifurcation of the system is delayed or occurred in advance.

With the development of robot technology and the advancement of intelligent processes, inspection robots are adopted for substation inspection instead of manual inspection, which reduces the workload of substation staff and improves the efficiency and quality of the inspection. The substation covers a large area and has a wide distribution of equipment. It is difficult for a single inspection robot to complete all the inspection work. At this time, multiple inspection robots are needed to cooperate by keeping a fixed formation. Aiming at the formation obstacle avoidance task of a multi-inspection robot system, this paper analyzes the force of the multi-inspection robot system by establishing a virtual potential field and designs the formation control algorithm for the robot system. Then a simulation system of the multi-inspection robot is designed based on SimMechanics and Simulink. Finally, the effectiveness of the proposed formation control algorithm is verified. The results show that the formation control algorithm designed in this paper can ensure that the robot system avoids obstacles to complete the target formation.

To mitigate the risk of transmitting dangerous bacteria and viruses in restaurant settings, this paper develops a restaurant intelligent beverage service robot with a low price but with good performance of gesture control. The robot is composed of a mechanical structure, a hardware system, and a software system. The first is the robot structure design, which includes the design of the beverage machine model and the design of the bottom wheel tray. Secondly, the hardware system includes the main controller, motor driver, visual camera, and sensor modules, and the ROS communication control system is used by a series of robot modules to communicate with each other to realize the restaurant area's autonomous path planning and accurate distribution of drinks. To address the issue of low gesture recognition rate in neural network algorithms, the system optimizes gesture frequency through cumulative recognition and judgment of software program calculation. This effectively reduces false recognition. Finally, through the ROS platform client experimental test, the robot motion control is stable and reliable, and the gesture recognition accuracy reaches 92.7%, which can meet the needs of the restaurant.

All input data are collected manually, which will not only bring a heavy workload to the operator, but also consume a lot of time. In order to quickly and accurately complete the robot calibration task, it is necessary to automatically complete the calibration of the entire system, so this paper studies a robot autonomous calibration method based on machine vision. By using the calibration board to establish the spherical coordinates and the camera field of view to plan the camera pose, you only need to place the calibration board within the measurable range of the camera to perform autonomous calibration. In the end, data collection time was reduced from over 10 minutes to less than 1 minute.

Given the increasing demand for robots with advanced environmental adaptability across various fields, this study aims to address the limitations of wheeled robots in traversing rugged terrains and the instability of legged robots on flat surfaces. In order to overcome these challenges, a hexapod robot with variable morphology was designed and developed. This robot possesses both flexibility and stability, enabling it to navigate a wide range of terrains. Through the measurement and analysis of the robot's motion states, the results of different motion scenarios were studied to validate the rationality and feasibility of the hexapod robot's structure. This research provides a foundation for further in-depth studies of this model.

The considerably increased number of high buildings with the urbanization development in recent decades has caused more difficulties for workers to clean the walls and glasses. Therefore, the wall-climbing robots developed rapidly because of their intelligent design and multi-functional cleaning performance. In this work, the wall-climbing robot is designed with atmospheric pressure as the attaching force to improve cleaning efficiency. The innovative design of the mechanical feet composed four main parts, such as the sucker, spring, supporting structure and force-detecting section. The typical adsorption structure is applicable on most wall surfaces especially the inclined surfaces with vertical or larger inclination angles, which significantly enhances cleaning efficiency and safety.

To address the need for robots to control external forces at their ends when performing shaft hole assembly tasks, this paper proposes a robot compliant assembly method based on six dimensional force sensors, and uses this method to perform operations such as hole searching and inserting. Firstly, the D-H method is used to define the coordinate systems of the robot arm, and its kinematic equations are derived, providing a theoretical basis for the following hole searching strategies; Then, based on the UR5 collaborative robot, an experimental platform for shaft hole assembly is built, and a six dimensional force sensor is installed at the end of the robot arm to achieve its end external force sensing function; Secondly, aiming at the problem of hole positioning, a spiral trajectory hole searching strategy based on force perception is proposed for fast hole searching operations; Finally, an experimental scheme was designed to validate the drilling strategy through experiments, and the experimental results were analyzed to verify the effectiveness of the method.

Aiming at the complex micro-device assembly robot system that has been built, based on the research on the composition of the motion mechanism and the micro-vision system, and the research on the micro-vision imaging system, derived image Jacobian matrix for parts assembly control based on single-channel or multi-channel micro vision, and obtained the conversion relationship between the pose change of the motion mechanism coordinate system and the micro-vision. And through the method of active movement of each movement mechanism, the image Jacobian matrix calibration between the macro-moving robot, the parts adjustment table, the micro-moving module and the micro-vision is realized. A part detection and adjustment strategy are formulated to realize rapid and effective control of micro parts. Experimental results verify the correctness and accuracy of the calibration method.

Public restroom cleaning robot has broad commercial application prospects, but there are still some challenges in practical application. Especially for the cleaning tasks in areas with stubborn stains such as washbasins and urinals in restrooms, the complexity is high, and the requirements for the flexibility of robots in the cleaning process are also strict. It is difficult to meet the requirements using traditional cleaning mechanism design. To address this issue, a design scheme for an intelligent sweeping, dragging, and cleaning integrated robot based on redundant dual arms is proposed; The designed cleaning module based on a dual arm robot is located at the front of the robot. By clamping the spray gun, cleaning operations are carried out on areas with stubborn stains. By switching between clean water and cleaning solution for continuous spraying, further deep cleaning is achieved; Then, the dirt collection and storage are completed by the collection device and the closing structure. While completing the three-dimensional structure design, based on the robot's motion posture and path planning, the urdf rigid body model was established, kinematics and work spatial analysis were carried out, and the weighted path planning was completed. The completed simulation results indicate that the robot has good motion stability and will not experience jumps during the motion process; Weighted path planning demonstrates the feasibility of robot cleaning with high coverage. Finally, we verified the feasibility of the toilet cleaning robot in kinematics and path planning, providing a feasible idea for the follow-up improvement of the toilet cleaning robot.

Operational cleaning robot face numerous challenges when attempting to deftly and steadily grasp objects in cluttered scenes due to factors such as limited areas, stacked items, and restricted sensor perception. To address this issue, we propose CR-Graspnet, a six-degree-of-freedom (6-DoF) grasp generation network that utilizes point cloud contact features. This approach decouples the grasp pose in high-dimensional space by defining contact points, allowing for joint learning of contact point sampling, grasp parameter regression, and grasp quality classification. Our experimental results demonstrate the effectiveness and feasibility of this method, with a success rate of 93% achieved in single target grasping scenarios.

In the multi-layer multi-pass welding of intersecting pipe fittings, the traditional teaching method is labor-intensive and has low welding accuracy. To weld more intelligently and accurately, this paper proposes a robotic multi-layer multi-pass welding filling strategy, determines the layout points of the weld passes, and plans the pose of the welding torch. In addition, this paper gives the transformation relationship between the welding seam coordinate system and the welding torch coordinate system, and finally simulates the weld passes.

In order to meet the requirements of technological development and higher exploration tasks, and to improve the applicability of underwater robots, this paper combines the advantages and characteristics of remotely operated submersible and autonomous underwater vehicle, and researches and develops a new hybrid drive underwater robot, which has two modes of thruster drive and cross rudder control, with the advantages of both full drive and under drive. In full drive mode, it has the advantages of fast navigation speed and good maneuverability; in under drive mode, it has the advantages of low energy consumption and large range. Compared with conventional underwater robots, the motion of hybrid-driven underwater robots is more complex and therefore places higher demands on the system design of the robot. In this paper, a system solution design for an all-drive under-drive hybrid underwater robot is proposed, assembled in real life and tested in the water. Through validation, the feasibility of the designed solution is demonstrated, as well as achieving the goal of improving the applicability of the underwater robot.

Compared to flat ground, objects in indoor environments have various curved surfaces, cleaning these surfaces is a higher-level task than traditional ground cleaning. To accomplish this task, we propose a new method to perform wiping on the curved surface of an object and complete the cleaning work through the end tool of the robotic arm. Our method is divided into two parts: An RGB-D semantic segmentation attention-based Feature Fusion Network (AFFNet), which can effectively fuse the features from the encoder and decoder to improve semantic segmentation accuracy; A path planning algorithm based on point cloud, which can autonomously generate the robotic arm operation path. The experimental results show that AFFNet achieves an mIOU accuracy of 46.24% on SUNRGBD dataset with excellent performance, the robotic arm can complete the curved surface cleaning operation with continuous and smooth path.

In this study, a soft robot with dual movement modes was designed to address the problem of pneumatic soft robots having a “tail,” limited movement, and a single movement mode. Inspired by the movement pattern of the catapult and the crawling gait of the tortoise, the design of the soft robot with dual movement modes was completed, including the crawling mechanism, jumping mechanism, and overall mechanism. For the crawling mechanism, a mathematical model of the deformation of the crawling drive was established using the Yeoh model and the principle of virtual work, and the factors affecting the deformation of the crawling drive were analyzed. The deformation of the crawling drive was simulated and analyzed using ABAQUS, and the relationship between air pressure, size of the drive, and deformation of the drive was investigated. For the jumping mechanism, the energy storage mechanism was analyzed in terms of energy storage and drive air pressure, and ADAMS was used to explore the jumping performance of the soft robot under different energy storage conditions for simulation analysis. The results revealed that the proposed design allows the robot to crawl and jump at the same time and removes the “tail” of the pneumatic soft robot, thereby improving its range of motion.

With the development and wide application of advanced weapons, the battlefield environment has gradually become more complex and harsher, which greatly increases the risk of rescuers performing search and rescue tasks. In view of this situation, this paper proposes a six-wheeled portable reconfigurable robot, Antibot, which is used to replace rescuers to perform life detection tasks in complex and unknown environments. This robot can adapt to the terrain through its unique three-rocker-leg passive suspension, and can be folded up for users to carry. The terrain adaptability of the robot is analyzed through mathematical modelling and numerical simulations, and verified by outdoor experiments. This robot has the advantages of simple structure and strong terrain adaptability, which has broad application prospects in battlefield rescue missions.

As the aging of the population becomes increasingly prominent, the health risks closely related to it should be paid attention to and prevented in time. Based on researching the conditions of lower limb rehabilitation training, and analyzing the basic structure and motion mechanism of human lower limb rehabilitation training, this article proposes a design scheme of a robot for lower limb rehabilitation training. The structure design of the lower limb rehabilitation training robot with tandem and hybrid sitting posture is completed, and the kinematic model of the lower limb rehabilitation training robot is established. The rehabilitation robot can realize the pitch movement of the foot and the adjustment of the leg rehabilitation tandem mechanism, with high flexibility and universality. It can meet the needs of different patients and help patients to carry out lower limb rehabilitation training.

In the realm of intelligent agent control, path planning has been one of the most well-liked study subjects. The path planning for obstacle avoidance approach proposed in this study is an enhanced artificial potential field method. By directing the barrier in the robot's direction and combining the boundary repulsion and vector decomposition of the goal direction, the local minimum problem is resolved. Next, the final combined force is delivered to the robot's center and the angle coefficient is added to provide a fair obstacle avoidance effect while taking into account the physical distance between the two sides of the robot and the quantity of extra environmental obstacles. Finally, the enhanced algorithm is used to perform a global path search. The simulation results validate the applicability and effectiveness of the suggested approach.

For robots, grasping is a necessary skill, and agile grasping by robotic hands is currently a cutting-edge research hotspot. Tactile perception serves as the primary sensory channel for both humans and robots to perceive the surface properties of objects, playing a fundamental role in enabling robots to perform dexterous operations. Tactile perception allows a robotic arm to sense the geometric contour, roughness, hardness, and other material properties of an object during the grasping process. It provides the robotic arm with predictive information on force and angles, thereby enhancing the efficiency of grasping. In recent years, deep learning has achieved remarkable advancements in various industrial domains, including intelligent sorting, defect detection, textile manufacturing, and autonomous driving. These achievements have spurred researchers to shift their focus from machine learning to deep learning in the study of tactile perception for agile robotic manipulation. Despite the unprecedented progress made in deep learning-assisted robot tactile perception, there are still some unresolved challenges in this field. This article begins by discussing the implementation methods of robot tactile perception and then provides a comprehensive overview of the current research and application status of deep learning-based robot tactile perception. Firstly, it highlights the latest advancements in tactile perception for dexterous robot operations. Secondly, it presents an overview of the sources and data acquisition methods employed in existing research. Additionally, the article summarizes the applications of deep learning in robot tactile perception. Finally, it explores current trends and potential future research directions in the field of tactile perception during robot grasping.

In order to solve the problem that the detection performance of the radar target detector decreases badly in non-homogeneous environments. Based on the actual echo clutter distribution, a Multiple Switching CFAR detector (MS-CFAR) is proposed. Firstly, the maximum reference units in the left and right reference sub windows are removed by censoring threshold. Then, compare the number of remaining reference cells in the left and right reference windows and variation index with the corresponding threshold respectively, and select the appropriate reference cells to accurately estimate the background noise power. Compared with the simulation and analysis results of other detectors, the proposed detector had the best detection performance and stability in multi-interfering targets, clutter edge and other non-homogenous environments. The results show that the proposed detector still has a good detection performance in non-homogeneous environments.

3-axis force sensors for both normal and shear force sensing have a promising prospect in the field of robotic applications. In this article, a novel flexible 3-axis force sensor was fabricated. The sensor is featured by a three-capacitor architecture composed of two asymmetric electrode levels separated by a microcone structured dielectric layer. The sensor showed a favorable normal force sensitivity of 2 kPa^-1 in the range of 0-100 Pa and a highly linear response to shear force in the range of 0-0.5 N with a sensitivity of 0.063 ^N-1. The sensor also demonstrates reliable sensing performance on both planar and curved states in cyclic test.

In order to effectively accomplish the control of robot in unknown environments, we propose a soft actor-critic- ResNet(SAC-ResNet) model based on multi-sensor information. First of all, comprehensively consider issues such as obstacle avoidance and robot walking efficiency, we design a state space based on various sensors and a continuous action space. Secondly, SAC combined with ResNet, multilayer perceptron(MLP) and multi-head-attention realize the organic integration of multi-sensor information, effectively solving the problems of low efficiency and environmental dependence caused by insufficient environmental information. The simulation results show that the SAC-ResNet model based on multi-sensor information can continuously control the robot to avoid unknown obstacles effectively in different environments, improving the accuracy, safety and robustness of local obstacle avoidance.

In order to analyze the welding trajectory of the rear axle weld on a drive axle, the kinematics equation of the robot was established using the D-H parameter method. MATLAB Robotics Toolbox was utilized to verify the forward and inverse kinematics through simulation and to plan the trajectory, continuous and stable terminal pose, change curves of Angle, angular velocity and angular acceleration were obtained. By incorporating specific weld points on the workpiece, the fitting toolbox in MATLAB was employed for curve fitting. The results demonstrate a strong alignment between the trajectory curve fitted by the actual welding points and the simulated trajectory curve. It provides theoretical support for the further research of MOTOMAN-MA1440 robot.

The location and perception systems are important component of indoor mobile robots, and research on their failure mode and effect can provide guidance for improving the performance of indoor mobile robots. In order to solve the problem that the weight distribution of risk factors in traditional failure mode and effect analysis (FMEA) is difficult and the evaluation results are greatly affected by subjectivity, an improved FMEA method of mobile robot location and perception systems based on interval triangular fuzzy number is proposed. Firstly, expert weights are allocated by multi standard analysis method, then the weights of risk factors are determined based on fuzzy analytic hierarchy process (FAHP), and finally the failure modes are ranked based on grey relational analysis (GRA) method. The experimental results indicate that the method proposed in this paper is more reasonable in risk ranking and can clearly distinguish between primary and secondary risks.

This paper presents a human intelligent editable program robot based on aelos platform. The robot uses artificial intelligence technology and features editable programs that allow users to program and customize them according to their own needs. Through the analysis of the Aelos Blockly platform, this paper proves that this robot is highly flexible and applicable, and can be applied to many fields, such as home, education, entertainment, etc. This paper also introduces the design principle and technical details of the robot, and gives some application examples to prove its practical effect. This study provides an important theoretical and practical basis for the development of the AI field, and also provides important enlightenment for the in-depth research and application of the AI technology in the future.

To meet the needs of the guidance of algorithm search in a complex environment, the optimality of global path in a static environment, and the security of real-time obstacle avoidance in a dynamic environment for path planning of cultural and tourism service robots, an algorithm based on the fusion of improved A * algorithm and Dynamic Window Approach is proposed. Firstly, based on the traditional A * algorithm, evaluation function are improved to improve the algorithm's search directionality to a certain extent; Secondly, the concept of safe distance is introduced, and a cubic broken line optimization method is proposed, which eliminates redundant nodes and inflection points, and only retains necessary key path points, greatly reducing the number of inflection points and improving the smoothness of the path. Subsequently, a dynamic obstacle vertical distance cost function was added to the evaluation function of the dynamic window approach to effectively reduce conflicts and collision risks between robots and dynamic obstacles. Finally, the improved A * algorithm is integrated with the dynamic window approach, selecting critical path points as temporary target points for the dynamic window approach. The dynamic window approach is used in segments for local real-time path correction, ensuring the optimal global path and avoiding collisions with unknown obstacles, ultimately achieving a safe and fast destination.

Aiming at cleaning requirements of stainless-steel wall surface for nuclear plant station (NPS) fuel pool, an underwater cleaning robot with two motion modes of swimming and crawling was proposed. According to the structural symmetry characteristic, the mathematical hydrodynamic models for two motion modes were established respectively. The robot was simulated with a Fluent based mathematical model to test its movement of direct navigation, rotary motion in swimming mode and forward navigation in crawling mode. The coefficient corresponding to the mathematical hydrodynamic model was got from the simulation. The results show that the second order hydrodynamic coefficient is dominant. At the swimming mode, the resistance of downward motion is much larger than that of the upward motion. The lift force of sway direction is close to surge direction, while the resistance is much larger. Compared to the surge direction at swimming mode, the lift force is much larger at the crawling mode while the resistance is little changed. The research results provide important data basis for structural design and motion control of the cleaning robot.

This study analyzes the impact of weld seam position and torch position on welding quality, including their definitions and spatial coordinate system representation. Based on this analysis, a coherent weld seam characteristic coordinate system is established, and the calculation of torch position is completed. This lays the foundation for further research on the inverse kinematics analysis of the arc welding robot. The plug-in arc welding robot is analyzed using forward and inverse kinematic analysis based on the general industrial robot. The inverse kinematic solution of the robot is then determined using the geometric method according to the specific coherent welding trajectory of the plug-in arc welding robot and its specific operation mode. The kinematic inverse solution of each joint is unique, the angle changes periodically, and its attitude meets the welding requirements of the torch, while the displacement changes of each joint can meet the working needs in the task space.

Robots are more and more applied to the automation of fruit picking, for the picking of large fruit trees, 7-Dof robots with excellent obstacle avoidance ability have a useful place. However, the inverse kinematics solution of 7-Dof robot kinematics has the problem of low precision and many iterations. In order to solve these problems, Pelican Optimization Algorithm is introduced to optimize the redundant robot. Pelican optimization algorithm is a new meta-heuristic algorithm, which has high development ability and high space exploration ability, and is suitable for solving optimization problems. In this paper, 30 groups of positions in the motion space of 7-Dof robot are randomly selected, and the inverse kinematics is solved by Pelican optimization algorithm. The simulation results show that using Pelican optimization algorithm to solve the inverse kinematics of 7-Dof robot can basically reduce the solving error to less than 10^(-7) mm after 250 iterations, and the iterative calculation time is only about 0.26s, which has better comprehensive solving effect compared with other intelligent algorithms.

When underwater autonomous robots operate underwater, sonar technology plays an important role. Firstly, the problems active and passive sonar face when applied to underwater target detection and their solutions are introduced. Secondly, the multipath effect and Doppler effect caused by the transmission distance more significant than the water depth, underwater environmental noise, and the relative displacement of the transmitting and receiving nodes during the underwater sonar transmission are introduced, and the corresponding solutions are listed. Thirdly, the development and application of the two-stage detection RCNN series and single-stage detection YOLO series in object recognition are introduced. Finally, the advantages and disadvantages of the above methods are summarized and discussed as the future development direction prospects.

At present, most of bionic robotic fish in BCF mode make use of reciprocating positive and negative rotation of a servo to drive the swing motion of the tail fin. However, the swing frequency of the tail fin of the robotic fish is relatively low and the generated propulsive force is also relatively small due to the influence of the properties of the servo. The other robotic fish which use DC motor or other unidirectional rotation motor to drive the tail fin have to take advantage of complex mechanical structure to realize reciprocating swing motion of the tail fin, even so, the swing amplitude of the tail fin cannot be adjusted directly. This paper presents a novel bionic swing propulsion mechanism which is driven by a stepper motor. The tail fin can swing at a high frequency by controlling the rotate speed in positive rotation of the stepper motor, and correspondingly, the swing amplitude of the tail fin can be adjusted by controlling the stepper motor to reverse rotation. In addition, a steering mechanism is added to this bionic swing propulsion mechanism to realize the steering function for the propulsion mechanism. The presented swing propulsion mechanism can be directly used for a swing mechanism of tail fin of robotic fish, or even can act as a kind of bionic propeller of underwater vehicle to replace the traditional screw propeller. In order to verify the feasibility of this underwater bionic propulsion mechanism, a series of swimming experiment were carried out.

In recent years, long endurance middle-large size unmanned aerial vehicle (UAV) has made significant progress in the industry field. Battery electric power system is easy to control for UAV, it is facing the challenge of poor energy density. Due to the fuel energy power system's ability to offset the energy density problem, this paper proposes a 25kg class hybrid power fixed-wing vertical takeoff and landing (VTOL) UAV. This UAV offers advantages such as long endurance, low energy consumption, and high payload capacity. Ground experiments and prototype flight tests were conducted to evaluate the performance and feasibility of the hybrid power system during hover flight. The results of these experiments demonstrated that the UAV achieved a flight time of up to 20 minutes during hover operations, compared to the 5 minutes of flight time for the battery electric compound-wing VTOL UAV. This experiment confirms that the hybrid power system enables the realization of a long-endurance compound-wing VTOL UAV

A novel driving principle of electromagnetic driven Spherical robot is proposed, combining the two principles of changing the center of mass and conservation of angular momentum, and a prototype of electromagnetic driven Spherical robot is developed. In this paper, the prototype is taken as the research object, and the dynamic analysis is carried out on its climbing ability. To improve the dynamic performance of the spherical robot, the influencing factors of the dynamic performance of the spherical robot are studied: one is the ratio of the equivalent pendulum mass m of the spherical robot to the total mass (M+m) of the spherical robot, and the other is the length of the equivalent pendulum of the spherical robot the ratio of r to the radius R of the spherical robot. Then the maximum climbing angle model of the spherical robot is deduced. The experimental results show that the design of the spherical robot is reasonable and the dynamic analysis is effective.

This paper proposes a double-ring clutch-type electromagnetic drive spherical robot. By controlling the position of the center of mass and the rolling direction of the left and right hemispheres, the omnidirectional rolling of the spherical robot can be realized, and the turning motion of the robot can be realized by using the principle of hemisphere differential. The left and right hemispheres of the clutch-type electromagnetically driven spherical robot have two states of unfolding and closing. When the two hemispheres are in the unfolded state, it has the characteristics of a wheeled robot and can move forward or backward smoothly and quickly in a better terrain environment. When the left and right hemispheres are closed, the double-ring clutch electromagnetically driven spherical robot has the advantages of "not afraid of falling over and flexible movement" of the spherical robot. The simulation research based on ADAMS technology shows that the stability of the spherical robot is poor when it is in the closed state, but it has good stability when it is in the unfolded state. Therefore, the correctness of the driving principle and the rationality of the structural design of the hemispherical differential spherical robot are verified.

In order to ensure that the robot has good mechanical properties, stability and hydrodynamic characteristics in underwater motion, and to protect the ROV core control module, battery compartment and other internal precision systems, this paper develops and designs a streamlined underwater robot shell, operating at a depth of 50m, with an overall theoretical linear design, regular shape using strength theory calculations and stress simulation analysis methods, and irregular shape using Solidworks Simutian numerical simulation method, based on Fluent for the overall shell in the fluid stress analysis. The results show that the stress σ₁= 9.33MPa and yield pressure p_CR1 = 13.368MPa for the large spherical shell; the stress σ₂ = 6.67MPa and yield pressure p_CR2 = 13.44MPa for the small spherical shell, both of which are less than the theoretical stress of 42.5MPa and yield pressure P_j=0.8MPa to meet the design requirements for strength and stability; the circular table shape and parabolic The maximum equivalent stress is 0.728MPa ≤ σ_s=50MPa, which meets the physical properties of ABS material; by simulating the stress changes in the three degrees of freedom of the ROV shell in and out, lift and rudder, 527.202Pa, 535.1Pa and 529.1Pa respectively, The yield strength of ABS material is 50MPa, which is much higher than the stresses in the three degrees of freedom of the ROV shell inlet and outlet, dive and rudder, indicating that the ROV meets the stress design requirements in the water.

In traditional underwater robot anti-submarine research, it is difficult to accurately estimate the detection performance of the underwater robot because the avoidance behavior of the underwater target is not considered, which makes it difficult to plan and optimize the working depth and other parameters of the underwater robot in advance. In this paper, the game theory framework, which is widely used in the study of economic, social, political and biological phenomena, is introduced into the research of depth optimization of anti-submarine work of underwater vehicles. The framework considers the hydrologic environment, moving path and acoustic characteristics of underwater target, introduces the concept of game between underwater robot and underwater target, and transforms the detection problem into a game problem. The game problem of maximizing one side and minimizing the other side is a two-person zero-sum game, and there is a mixed strategy Nash equilibrium solution in the game between the underwater robot and the underwater target. The validity of this analysis method is verified by a concrete case.

The classic Kalman filter (KF) seeks to reduce the variance of the estimation error; however, owing to the uncertainty of the system model and the statistical properties of noise, its application is restricted. The mentioned issues can be efficiently fixed using the H_∞ filter. The connection between the Krein space KF and the H_∞ filter is established, and the necessary and sufficient conditions for that are the ܪஶ state and disturbance joint one-step smoothing (SDJOS) estimator, which is determined based on the definition of the H_∞ performance index and state transition matrix (STM) of the linear discrete-time fractional-order system (LDFS). The H∞ SDJOS estimator based on the piecewise Riccati equation is provided using the innovation analysis technique. An example is given to demonstrate the effectiveness of the suggested approach.

In order to improve the absolute positioning accuracy of the robot, a robot -based robot zero and hand -eye integration calibration method is proposed. By collecting several robotic joint corners, end position data data, and cooperative target position data under the camera coordinate system, use the string function to adjust the input, and then based on the gradient drop method, obtain the results of the robot's eye and eye calibration. Errors and correction of their parameters, thereby improving the absolute positioning accuracy of the robot. The experimental results showed that the maximum error of the absolute positioning accuracy of the robot was reduced from 6.298mm to 3.149mm, and the average error dropped from 3.745mm to 1.899mm. It indicates that this method can improve the absolute positioning accuracy of the robot under low cost conditions.

Despite multi-branches network can strengthen the ability of feature extraction, the cost is speed reduction. Thus, in order to improve FaceBoxes’ accuracy and do not reduce speed, we propose Reparameterization-FaceBoxes in this paper that its multi-branches modules are replaced with diverse branch block. We make comparative experiments on WIDERFACE dataset, Reparameterization-FaceBoxes outperforms FaceBoxes with higher accuracy and faster speed, it can increase by 0.9%, 0.8% and 0.4% on Easy, Medium and Hard datasets while its speed increases about 20% for VGA-resolution images.

In zinc flotation process, concentrate grade is an important indicator which cannot be measured online. To ensure the stability of froth flotation process, the deep learning has been widely used for the prediction of concentrate grade. However, the grade prediction based on deep learning leads to the dataset bias problem. A popular network for addressing dataset bias issues is the domain adversarial neural network (DANN), but the domain prediction of DANN is a binary classification, which cannot accurately reflect the data distribution of dataset. To solve the deficiency of the binary classification, a wasserstein domain adversarial neural networks (W-DANN) is proposed, which calculates the domain loss with wasserstein distance and enhances the domain prediction performance. Finally, the experimental results on the Office31 dataset and flotation dataset demonstrate the effectiveness of the proposed W-DANN model and its possibility of application in zinc flotation process.

In this work, we propose MotionFormer, a multi-object tracking method incorporating motion information. MotionFormer is a Transformer-based architecture, which is an attention-based query-key mechanism. A dynamic Kalman filter is proposed to predict the position when the object is occluded and the appearance characteristics are destroyed. TIoU (Tracking Intersection over Union) uses motion to improve the IoU distance measure, making it possible to focus on effective location information around objects. On MOT17 benchmark, MotionFormer achieves 74.9% MOTA.

This paper addresses the decentralized simultaneous multi-agent task allocation problem in saturation attack scenarios. This problem arises when multiple agents have to cooperate simultaneously to perform complex and dangerous tasks that may result in their malfunction or damage. A modified consensus-based grouping algorithm with virtual tasks (CBGA-VT) is proposed that considers time windows, different types of tasks, and destroyed agents after performing certain tasks. Virtual tasks are introduced to address the decreased agents problem and the scoring scheme is improved to achieve better performance. Simulation results demonstrate that CBGA-VT can effectively solve the multi-agent simultaneous task allocation considering destroyed agents in saturation attack scenarios.

Unmanned Aerial Vehicles (UAVs) are increasingly used in a wide range of applications, and the issues of autonomous landing and endurance are gaining greater prominence. Image processing methods have been attracting attention on account of their edges in high accuracy and low cost. Firstly, the overall scheme and workflow of the UAV autonomous mission system are introduced in this research; DJI M100 UAV is selected as the research object, and STM32F407 is selected as the airborne control board; then this paper focuses on the precise positioning technology of UAVs based on image processing, introduces the UAV image extraction algorithm, processes the extracted image with the maximum between-class variance method (OTSU), and transforms the points in image space into Hough space by Hough transform based on Hough line detection algorithm, in an attempt to find out the lines in image space and determine the UAV position; consequently, the precise positioning device of the UAV is designed, and the battery is replaced independently to complete the endurance function design; finally, through the field test experiment of UAV, the results demonstrate that the image processing detection method boasts high accuracy and good detection accuracy. It is available for UAVs to automatically change batteries to achieve autonomous battery life and high stability.

In this paper, based on adaptive control and terminal sliding mode control method, we introduce a new four-dimensional fractional-order (4-D FO) Rösslor hyperchaotic system with uncertainty and external disturbances, and design a new controller to study its synchronization problem. First, a fractional order sliding surface is designed to make the error dynamics converge to near zero in finite time on the sliding surface, and then, a new sliding controller with adaptive updating laws is designed , which eliminates the effect of nonlinear terms in the systems and makes the error system converge to zero quickly , this results in synchronized of the drive and response systems, finally, based on the 4-D FO Rösslor chaotic system, numerical simulations show that the controller based on adaptive control and terminal sliding mode designed in this paper has good practicality and feasibility.

Ant Colony Optimization (ACO) is a classic swarm intelligence optimization algorithm that has been widely applied in various task scheduling scenarios. However, traditional ACO may easily get trapped in local optimal solutions. Inspired by Hybrid Breeding Optimization (HBO) algorithm and coevolution, this paper proposes a Heterogeneous Coevolution Ant Colony Optimization (HCEACO) algorithm based on hybrid breeding mechanisms to overcome the shortcomings of a single population in terms of solution diversity. Moreover, a strategy based on population similarity is proposed to determine whether communication is necessary after a fixed number of iterations, and to maintain a dynamic balance between population diversity and convergence speed in selecting communication partners. To fully validate the effectiveness of the proposed algorithm, multiple path planning algorithms are simulated and applied to multi-load Automatic Guided Vehicle (AGV) path planning. The experimental results show that the improved algorithm performs well in the multi-load AGV path planning problem, and has broad application prospects in this field.

The Radio-Determination Satellite Service (RDSS) is a unique and advantageous service of the Beidou system, and aircraft equipped with Beidou RDSS can be widely used in rescue sites. This paper evaluates the security of ground 2.5GHz 5G base stations and aircraft with Beidou RDSS systems. In the study, the dynamic simulation results of the impact of different flight altitudes and horizontal isolation distances on the system performance in urban and rural nonline of sight scenarios are given through aggregate interference simulations. The analysis results show that the higher the flight altitude of the aircraft, the easier it is for the two systems to coexist, and the rural scenario is easier to coexist than the urban scenario. Finally, the conditions for the safe operation of the 2.5GHZ 5G system and the RDSS aircraft are derived to provide a reference for practical situations.

Aiming to address the issue of pose detection of workpieces in automatic assembly processes, this paper presents an algorithm for detecting the pose of metal gear workpieces. The method utilizes an improved homomorphic filtering algorithm to eliminate spot noise from metal workpieces. The complete gear profile is then obtained through the use of the K-means threshold segmentation method and neighborhood tracking method. The center of the gear is determined with the gravity center method, and a virtual circle is scanned with the center as its point of origin. The subpixel feature points of the tooth profile are determined by the least square method, enabling the exact position and pose of the gear workpiece to be obtained. The results of experimentation demonstrate that this method improves detection accuracy by 41.7%, reduces detection error, and has stronger robustness than pixel-level detection algorithms. The proposed method enables the high precision positioning of the position and pose of the gear workpiece and has practical application value.

Human society is a comprehensive group structure. As a special group, blind people occupy a large proportion in society. It is inevitable and very difficult for blind people to walk alone in their life. First, they are easy to lose their way at intersections. Second, some imperceptible obstacles will hurt them. Finally, the Blind Way occupied will make their walk more difficult. However, with the progress of computer technology, there are more and more products that can help the blind to travel, such as wearing the obstacle warning clothing on the body of the blind, or the obstacle recognition device in the form of glasses. However, most of the products are not only very complex in use, but also change the original travel habits of the blind, making them have a kind of resistance. This research is based on the concept of human-centred design from the perspective of helping the blind, with the focus on not changing the original habits of the blind, using the concept of barrier-free design as a guideline to discover the travel needs of the blind in their daily lives, and using the sensors as the collection end, and using the computer and the device for interaction, in order to complete the design of mobility products for the blind. It is hoped that the design and research in this paper will help the development of guide products and serve as a reference for future research on such products, and that more designers will pay attention to these visually impaired people and give them more and better designs in the future.

In the cell injection task conducted in the microgravity science glovebox of the space station, the ground operator remotely controls the space station's robotic arm to perform gene injection or nuclear extraction tasks. Reliable force feedback plays a crucial role in adjusting the needle's position and speed during the manipulator's operation to prevent puncture failure. This paper proposes a SPH(smooth particle hydrodynamics) method based on a unified particle model to simulate the interaction between the rigid injection needle and the fluid cytoplasm. A fast neighborhood particle search algorithm is utilized to obtain collision information between rigid bodies and fluids, and a cell injection interaction force model is designed to simulate the adhesion phenomenon between the injection needle and cytoplasm, which includes viscous, pressure, friction, and buoyancy forces. In conclusion, two sets of simulation experiments were designed to demonstrate the real-time stability of the method, as well as its effectiveness for force interaction simulations of highly viscous fluids.

Due to the limited load capacity of small unmanned aerial vehicles (UAVs), micro pressure sensors are generally selected to measure altitude and airspeed. However, micro pressure sensors have relatively large absolute error for altitude measurement when UAVs take off or land. Therefore, in this paper, an ultrasonic module is added to make up for the lack of altitude measurement when UAVs take off or land (altitude is below 10 m). This paper mainly introduced the principle of ultrasonic ranging and tested the validity of ultrasonic ranging data. The test results showed that for some complex situations, ultrasonic ranging data needs to be filtered, and it also showed that Kalman filtering could meet the requirements of unmanned ultrasonic low altitude measurement. Finally, this paper proposed the basic architecture of the UAV height fixing control system in combination with ultrasonic ranging technology.

In the current power system, the existing industrial gateways suffer from drawbacks including sluggish sampling rates, inadequate precision in data acquisition, and fragmented data collection points. To mitigate the excessive burden on production lines resulting from increased interaction caused by the aforementioned issues, this paper presents an advanced data acquisition model for industrial gateways. Based on edge computing and MapReduce, this model significantly improves the efficiency and performance of data collection from industrial production line points. By enhancing the gateway's data acquisition component and ensuring the security of gateway data, our proposed approach outperforms existing industrial gateways.

Visual acuity (VA) assessment is a critical screening method for evaluating ocular function and identifying ophthalmic disorders such as myopia. However, conventional VA tests reliant on human-human interaction suffer from limitations such as subjectivity, dependence on medical professionals, optotype memorization, and fixed subject-optotype distance. To address these limitations, this paper introduces an innovative Human-Machine Interaction (HMI) system specifically designed for VA assessment. The HMI system consists of a computer, an external camera, dedicated HMI software, and gesture input as its key components. The intelligent core of the system is the HMI software, which incorporates static gesture recognition, optotype size design, and a VA retrieval algorithm. The primary function of the HMI software is to process real-time data captured by the camera, dynamically adjust optotype sizes based on gesture recognition facilitated by the MediaPipe Hands module, and retrieve VA levels using the VA retrieval algorithm. Experimental results involving 50 participants demonstrate an average accuracy rate of 96.2% in gesture recognition, with specific accuracy rates ranging from 93.6% to 99.2% for different directions. The overall performance validates the HMI system's effectiveness in achieving highly accurate VA assessments.

An improved model combining convolutional neural network and bidirectional long and short-term memory network is proposed to realize the intelligent diagnosis of GIS device fault. First, the feature vector extraction of the speech data through wavelet packet decomposition, and the results are used as input to the CNN-BiLSTM model to learn the sound feature relationship of GIS devices. Secondly, it improves the ability of the model to intelligently diagnose faults of GIS devices by optimizing parameters and accelerating convergence. After experimental verification, the proposed model has obvious advantages in GIS equipment fault diagnosis over the traditional fault diagnosis methods.

The failure of air conditioning refrigeration system will lead to the rapid increase of temperature in data center, which will lead to server downtime. In this paper, the CFD method is used to simulate the nine working conditions of underfloor air conditioning, in-row air conditioning and overhead air conditioning, which are respectively composed of no containment, cold aisle containment and hot aisle containment. Taking the time required for the maximum inlet air temperature of the cabinet to rise to the limit temperature (45 ℃) for the normal operation of the server as the evaluation standard, the reliability of the air conditioning scheme after the failure of the refrigeration system is studied. The results show that the reliability of hot aisle containment is the highest when using underfloor air conditioning and overhead air conditioning, and the reliability of cold aisle containment is as same as hot aisle containment when using in-row air conditioning scheme. In the case of hot aisle containment, the ability of the three air conditioning schemes to bear the failure risk of refrigeration system is as follows: underfloor air conditioning (> 900s) > overhead air conditioning (870s) > in-row air conditioning (570s).

The measurement data obtained by lidar in front of the transparent glass will have large errors with the actual environment. it is necessary to solve the problem of environmental perception of mobile robots in transparent glass indoors. This paper introduces four laser SLAM algorithms, Gmapping-SLAM, Hector SLAM, Karto-SLAM and Cartographer SLAM, and makes experimental comparison and analysis through simulation platform. The most suitable algorithm for indoor glass scenes is selected.

Aiming at the problems of high manual detection cost, too strong subjectivity and low level of production automation, an online detection algorithm for the crystallization status of penicillin salt solution is proposed. Firstly, on the basis of penicillin salt solution image processing, in order to eliminate the foam influence generated during the crystallization process, the color space is transformed and the three features of the penicillin salt solution image are extracted, secondly, the three features are used as evidence and the basic probability distribution function (BPA) of DS evidence theory is constructed, and the results are obtained by fusion according to the DS evidence theory rules, and finally, due to the irreversibility of industry, a constant false alarm detection based on Pearson's criterion is introduced to further improve the accuracy of penicillin salt solution crystallization state alarm. This algorithm is used for process control. The experimental results show that the online detection algorithm of the state of penicillin salt solution based on DS evidence theory has a fusion single detection probability of more than 85% under the condition of ensuring that the constant false alarm rate is 6 10, , and the online operation is consistent with the manual detection results, but it has real-time performance that cannot be matched by artificiality, which greatly improves the crystallization quality of the product.

Visual and linguistic modalities provide complementary information for various computer vision tasks. This paper proposes a novel approach for improving the tracking algorithm SiamBAN by integrating visual and language modalities. We introduce the concept of a visual-language modal mixer, which combines visual features and language representations to enhance the tracking performance. Specifically, we leverage a language model to extract semantic features from language descriptions and align them with visual features using a linear layer. The VL Modal Mixer is implemented through the Hadamard product operator, preserving spatial information. The mixed features are then fused with visual features through a residual connection to retain fine-grained visual details. Extensive experiments on benchmark datasets demonstrate the effectiveness of our proposed method, achieving state-of-the-art performance in terms of accuracy and robustness. Our work contributes to the advancement of multimodal tracking algorithms and opens up new possibilities for integrating visual and linguistic cues in computer vision tasks.

Face cartoon style transfer to generate high quality face images has become an art form pursued by many people, but face style transfer generally has problems of incomplete detail information after migration and poor generation quality in some exaggerated styles. In this paper, By proposing a network model based on StyleGAN suitable for face style transfer by improving the StyleGAN generator part, introducing a style restriction module to characterize the color as well as more detailed information, and the overall process uses a progressive image generation strategy to gradually generate highquality style transfer result maps. The results show that the method can not only achieve style transfer in both domains, but also reconstruct low-resolution images to better characterize their features with better visual effects.

There are two main challenges faced by generative adversarial networks (GANs) for facial attribute editing. The first challenge is how to perform targeted attribute editing in a controllable manner during the editing process, preserving the variability of relevant attributes while ensuring the invariance of irrelevant attributes. The second challenge is the high computational resource requirements of GANs, making them demanding on hardware performance for almost all image editing GANs. To address these challenges, we propose our SLGAN model based on StyleGAN2, which incorporates separable loss and knowledge distillation methods. Experimental results demonstrate that our proposed model achieves promising performance on relevant tasks.

This paper presents a series of enhancements made to the YOLOv5 model, which belongs to the well-established YOLO (You Only Look Once) series of object detection models. The proposed modifications yield a significantly advanced object detector exhibiting exceptional performance across diverse datasets. The primary focus of our improvements lies in the replacement of select convolutions within the model using an enhanced reparameterization technique tailored for convolutional models. In conjunction with other effective enhancement strategies, the augmented YOLOv5n model achieves a mean average precision (mAP) of 77.8 on the VOC2007 dataset, showcasing an impressive 18% performance gain over the original model (version 6.0). This notable improvement positions YOLOv5n ahead of the state-of-the-art YOLOv8 model, while concurrently attaining further enhancements in frames per second (FPS) compared to its predecessor. A comprehensive set of experimental results substantiates the efficacy of our approach towards enhancing the YOLOv5 model, rendering it more amenable to the requirements posed by various application domains within the field of object detection.

In response to the limited amount of cell detection data and the phenomenon of overlapping and adhesion in the existing bronchoalveolar lavage fluid (BALF), a YOLOv5s-based target detection algorithm (YOLOv5s-GAN) was proposed to optimize the model structure and loss function and effectively address the overlapping and adhesion of BALF cells. The algorithm enhances the accuracy of cell detection in BALF by using a lightweight convolution GSconv module and a normalization-based attention mechanism (NAS) that does not require additional calculations or parameters. Additionally, the algorithm improves detection accuracy by using D-IoU to mitigate false detections due to overlapping cells. Experimental results showed that the algorithm achieved an average accuracy of 88.9% in detecting four types of BALF cells, which demonstrated its high practicality for cell detection in BALF.

A new multimodal feature fusion representation method is proposed to address the problem of low accuracy in gait recognition under single-modal conditions. The contour features and skeletal features are obtained using background subtraction and human pose estimator, respectively. The two types of features are fused in an embedded manner to generate contour-skeleton feature images. The resulting skeleton-contour sequences are horizontally segmented and processed individually. Finally, the extracted feature vectors are mapped to a metric space using fully connected layers. Furthermore, in the existing GaitPart algorithm, the triplet loss function is employed for distance metric learning during network training. However, it suffers from slow convergence and unstable performance. To enhance the discriminative ability of the network, we propose a joint loss function that combines the triplet loss and softmax loss during the training process. Finally, extensive experiments are conducted on the CASIA-B dataset to evaluate the effectiveness of our proposed approach.

In this paper, the threat assessment of single UAV target is carried out according to the requirements of anti-UAV cluster operation. Firstly, the single target threat assessment framework is described, and the threat elements of UAV targets in a given environment are extracted, including static threat elements and dynamic threat elements, which are quantified by scale method. Then the threat weight of each element is analyzed, and the final weight is determined. Secondly, in view of the problem of poor prediction accuracy of FWNN algorithm in UAV threat assessment, GA-FWNN algorithm is proposed. The GA-FWNN threat assessment model is constructed by improving the FWNN algorithm through two methods: increasing the number of samples driven by data and updating the optimization parameters by genetic algorithm, and it is applied in the UAV threat target assessment simulation. Finally, through the simulation of the combat scenario, it is proved that the single UAV target threat assessment based on GA-FWNN has better prediction ability and can accurately estimate the combat target threat.

Simultaneous Localization and Mapping (SLAM) is a research hotspot with increasing demand for indoor navigation and mapping. Single-line LiDAR is an effective sensor for indoor SLAM due to its low cost and high accuracy. However, its limited field of view makes it difficult to obtain comprehensive 3D point clouds. In this paper, we propose a single-line LiDAR indoor SLAM based on image interpretation. Our proposed system utilizes the Cartographer framework as its backbone and mainly consists of four components, including input sensor data, sematic detection, local SLAM and global SLAM. This approach effectively addresses the limitations of traditional single-line LiDAR SLAM methods that do not incorporate image interpretation and Cartographer map. Experimental results show that our proposed method achieves accurate and robust localization and mapping in indoor environments, and outperforms traditional single-line LiDAR SLAM methods without image interpretation.

Traffic prediction plays a vital role in urban transportation systems for effective traffic management, congestion mitigation, and resource allocation. Traditional approaches often overlook the heterogeneity and complexities of real-world traffic systems. In this paper, we propose a novel approach, Sparse Heterogeneous Grid Traffic Prediction with Cross-Adaptive Multi-Graph Attention, which leverages graph neural networks (GNNs) to capture the intricate dependencies among road segments within a sparse and heterogeneous grid framework. The proposed model incorporates cross-adaptive multi-graph attention mechanisms to adaptively capture the varying influences and correlations among different road segments. Real-world traffic datasets are used to evaluate the performance of the proposed model against baseline methods. The results demonstrate the superiority of our approach in terms of prediction accuracy, robustness, and adaptability. The findings from this study contribute to the advancement of intelligent transportation systems and pave the way for more efficient and sustainable urban transportation networks.