This PDF file contains the front matter associated with SPIE Proceedings Volume 12720, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

We propose an intent prediction-based industrial data flow scheme, which is mainly designed to input and parse the user's intent, and then get the user's intent more precisely and make corresponding policy matching for the user's intent. Aiming at the blockage of a large number of industrial data circulation in industrial scenarios, the solution combines the prior prediction knowledge of intent to predict the intent of industrial data and develop customized circulation schemes for different types of industrial data. Experimental results show that the proposed scheme can effectively improve the blockage of large-scale industrial data flow, determine the data control strategy flexibly and quickly, and further improve the intelligence of data sharing and circulation.

Scientific estimation and dynamic monitoring of regional nighttime light evolution in long time series is an important source of information for assessing regional human nighttime economic and social activities. Based on the nighttime light remote sensing data of the long time series from 1992 to 2013, this paper took the two sides of Helan Mountain as the research object, extracted the total amount of regional light by using TNLI index, and analyzed the temporal variation characteristics of nighttime light on both sides of Helan Mountain from time and space by studying the nighttime light changes. The results show that the nighttime light of Helan Mountain and its two sides increased year by year from 1992 to 2013, and increased sharply in 1997, while the variation of total nighttime light and nighttime light density tended to be gentle from 2002 to 2013. In the temporal and spatial evolution of nighttime light in Inner Mongolia and Ningxia, the total amount of nighttime light on both sides of Helan Mountain has been increasing, but the total amount of light has changed little since 2002. The above indicated that in order to protect the ecology of Helan Mountain, the regional development on both sides of Helan Mountain changed from rapid expansion to stable upward. The research results can provide data support for the ecological protection of Helan Mountain and the formulation of policies.

This paper analyzes the research progress and existing problems of digital industrialization at home and abroad, and proposes an architecture that integrates data, computing power, cloud and network. The architecture consists of three layers: data acquisition layer, edge layer and center layer. And it puts forward solutions to four core problems involved in digital industrialization: data grading lakehouse scheme for data storage, data-intelligence integration for data fusion, privacy computing-based scheme for data circulation and blockchain-based scheme for data transaction. At the same time, the technical maps of data, computing power, cloud and network are summarized and analyzed.

This paper discusses how the coupling capacitor of transmission line in high speed circuit is optimized. Take the 400G bit error tester as an example, the characteristic impedance of the transmission line before and after adding the coupling capacitor is simulated. In order to reduce impact on signal integrity caused by impedance mismatch, the treatment method of voiding is obtained through the impedance calculation formula. By comparing several different voiding scheme proposed by predecessors, on the basis of the previous work, the hollowing treatment is further optimized to study the effect of hollowing size on signal integrity. Then HFSS software is used for modeling and simulation, and the insertion loss and return loss under several different schemes are calculated. It is found that the size of the hollowing process will have a certain impact on the integrity of the signal, and to a certain extent, the larger the hollowing size, the smaller the insertion loss and return loss.

A Q-band low-profile FSS with controllable numbers of the band is proposed in this paper. Using the complementarity principle, the branched strip lines are complemented as slots, and the FSS with controllable numbers of the band is designed. Then its working mechanism is analyzed by using the equivalent circuit method. Finally, the model is solved by incidenting Floquet mode using CST. The simulation result is fitted by the equation derived from the equivalent circuit method and shows that there are four center frequencies of the bands (or i.e. four FTZs): 36 GHz, 44 GHz, 47.5 GHz, and 48.2 GHz, respectively. By designing the two and three slots, the controllability of the numbers of the bands has been verified. By analyzing the parameters p, p, of the unit of FSS, the controllability of the position of bands has been verified. In addition, the horizon size of the unit is x=2.24 mm, y=2.02 mm, and the profile is z=0.82 mm which is a small size design. It achieves the design purpose of miniaturization and low profile.

FinFET (Fin Field-Effect Transistor) technology is widely used in advanced transistor manufacturing process due to its lower leakage current, smaller short-channel effect, and lower power consumption. However, this special physical structure is very susceptible to manufacturing defects, which makes the FinFET memory prone to functional faults. Therefore, an efficient fault detection method is very important for the defect detection of FinFET memory. A detection algorithm called March FRD for FinFET memory faults is proposed in this paper. This paper summarizes the FinFET functional fault model, analyzes the corresponding coverage conditions, and verifies the correctness of the algorithm. March FRD performs well in detecting resistance dynamic faults, and its total fault coverage reaches 91.7%, which is significantly better than other existing March tests.

This letter analyzes the spectral and energy efficiency of an intelligent reflecting surface (IRS)-assisted the dual unmanned aerial vehicles (UAVs) in target detection with hardware support, where devices first harvest energy from a power station (PS) in the downlink (DL) and then transmit information to a data sink in the uplink (UL). However, in order to increase the spectral efficiency, most existing works on dual UAVs adopted the simplified linear energyharvesting methodology and also cannot guarantee the detection performance. To improve system property, we optimize the beamforming coefficients on the basis of improved semidefinite relaxation (SDR) to minimize the total transmit power subject to the signal-to-noise (SNR) of the receiving antenna reaches the standard. Moreover, we analyze the energy saving effect of optimized system. The simulation results show that with the increased number of elements in the IRS, the proposed method has better detection probability and higher energy efficiency.

With the development of satellite technology, satellite communication can realize reliable communication with long-distance and large-scale coverage. However, satellite resources are now relatively scarce. To address this feature, this paper focuses on designing a reasonable and efficient elastic network model and prioritizing tasks with Mobile Edge Computing (MEC) technology to maximize the use of satellite node resources while handling urgent tasks promptly. Task offloading is a key technology for MEC, and the required parameters for modeling and optimizing target designs can be obtained through the resource virtualization function in the architecture of elastic satellite network. In addition, users can make efficient offloading decisions among servers on this architecture. We focus on the optimal delay model for a multi-user multi-edge server scenario with priority, i.e., the Software-defined Multi-priority Task Offloading Model (SMTOM). The optimization objective of the scenario is then reduced to a Mixed Integer Nonlinear Programming (MINLP) problem, for which Deep Reinforcement Learning (DRL)-Based Dynamic Task Offloading (DDTO) algorithm is used to solve the task offloading problems in satellite scenarios. The DDTO algorithm improves the average delay performance by 81%, 93%, and 84% compared to the random offloading algorithm, the local offloading algorithm, and the greedy algorithm, respectively.

As the rapid development of digital currency over recent years, the stability of financial markets and monetary policy of the central bank has been greatly affected. There is great significance that The People's Bank of China continues to promote the e-CNY with the strong support of the government under the situation. As one of the important ways to achieve the application of e-CNY, the development and research of NFC hard wallet near-field payment advanced with significant value. This paper focuses on the NFC hard wallet technology combining NFC and e-CNY , firstly describes the relevant policies and the principles of e-CNY, the principles of NFC, the architecture and process of mobile phone NFC, the memory process of NFC hard wallet and the power consumption optimization, and then summarize the available technical solutions, lastly put forward specific improvement measures by comparing users' usage habits and come up beneficial suggestions with the promotion and popularization of e-CNY.

The in-vehicle network communication system and the in-vehicle network control system are independent and inseparable, and the network performance and control performance affect and restrict each other. The purpose of this paper is to study the optimal design of a network control system for railway vehicles based on artificial intelligence. This article is a project actually implemented relying on the railway vehicle network control system in M province. This paper selects the MVB network that conforms to the IEC61375-3 standard through the analysis and comparison of several mainstream communication networks in the rail transit industry. In the process of network topology design, the network system is divided into two network segments, the vehicle network and the train network, through the rational use of repeaters. There are a total of 160 CCU source ports and sink ports. In the process of designing the network control system, we have fully considered various redundancy design schemes, such as master control redundancy, line redundancy, key signal redundancy, hard cable signal redundancy, etc., to ensure the stability and reliability of the entire system. In order to achieve the design goal of subway vehicles mainly based on network control and supplemented by hard line control.

With the rapid development of the economy, communication and information services are also developing in a swift and rapid direction. Communication technology is the transmission and exchange of various forms of information, and remote network control is the reliance on better communication technology to provide powerful control of machines networked over long distances. Networked remote control is constantly evolving, and studying the application of communication technology in networked remote control can improve its application. This paper describes the working principles of communication systems and clarifies the basic concepts of network control from the perspective of remote network control. Various optimisation algorithms used for communication are further investigated. The study of communication technology in network remote control can continue to improve the scope of its application, while strengthening the analysis of network remote control and better analysing the application of communication technology in network remote control, thereby increasing its spatial application and promoting its continuous development. The research in this paper can improve the stability of network remote communication control and the balance of channel transmission, improve the performance of computer network remote control and improve computer communication services.

With the rapid development of China's economy and the rapid improvement of technology, the Internet of Things is gradually appearing in people's vision, the network is becoming an indispensable part of people's lives, and the use of computers, cell phones and other devices for information exchange is the most common and convenient in daily life and work. However, at this stage, the security of IoT information technology has been challenged to a certain extent, which is a big threat to China's economic development, social progress and citizens' property security. In this context, this paper discusses the key technologies in the information security of IoT. This paper first introduces the current situation of IoT information security, and then proposes the use of network diversity technology for IoT security protection, which involves the introduction of diversity technology and the role of code diversity technology for IoT security. This paper focuses on how to use diversity as a security technology to effectively protect the information security of IoT, which is of great importance to guarantee people's normal life and protect personal privacy from infringement.

Facing an increasingly complex electromagnetic environment, modern communication systems must adopt certain antiinterference technology when deploying system equipment and network to ensure the normal operation of wireless communication. Currently, interference recognition is the foundation and key link of anti-interference technology. Among them, the recognition accuracy and the dependence of the algorithm model on training data are challenges that need to be solved urgently. In this paper, a CNN-RNN joint network architecture combining residual network and LSTM network is proposed to recognize the interfering signals. The joint network architecture adopts the parallel combination of residual and LSTM network, where the time-frequency image data of signals is input to the residual network branches while the real part, imaginary part, and spectral amplitude data of signals are input to the LSTM network branches. After simulation verification, the interference recognition result of the joint network is significantly improved compared with the single network. Firstly, compared with the single LSTM network, even though the single LSTM network has reached a very high recognition accuracy, the recognition accuracy of the joint network is still about 1%∼2% higher. What’s more, compared with the single network, the interference noise ratio (INR) generalization ability of the joint network is obviously improved. After training the network with different INR distributions, the recognition accuracy can be maintained. Therefore, it’s not sensitive to the INR distribution of the training data, which can adapt to different distribution conditions of training data and reduces the dependence of the algorithm on training data.

In the new era, in the face of the future battlefield with informationization, three-dimensional combat and flexible maneuverability, all major military powers in the world are vigorously promoting the intelligent strategy of traditional firepower weapon equipment, and the militarization application of artificial intelligence technology is gradually becoming a research hotspot of military experts at home and abroad. Firepower weapon equipment is also gradually undergoing a transformation process from "mechanization and automation" to "informatization and intelligence". Under the new combat background, this paper starts from the category of artificial intelligence technology, analyzes the key technologies such as target recognition and automatic speech recognition based on image information in detail, and indepth discussion and prospect of artificial intelligence technology will vigorously promote the future technical changes of traditional firepower weapon equipment.

In this paper, a simple dual-band polarization conversion metasurface(PCM) is proposed. This polarization conversion metasurface can be used in 5G applications such as satellites and radar. The polarization conversion is achieved by etching trapezoidal slots on the upper surface of the metasurface unit, resulting in a change in the distribution of the surface current. To achieve a dual-band, an additional resonant frequency is excited by loading two metal vias and two slots on either side of the metal via. In addition, the tuning of the frequency and the control of the operating bandwidth can be achieved by changing the radius size of the meta via and the slot width of the metasurface unit. The simulation results show that the line polarization(LP) conversion to line polarization(LP) is achieved at 37.6~59.5 GHz and 63~75.6 GHz with the average polarization conversion ratio of 94% and 96% in the two bands, respectively, and the lowest polarization conversion ratio of 90%.

Aiming at the characteristics of the long delay, nonlinearity, and large inertia in the heat exchanger temperature control system, combined with an intelligent algorithm and PID control system, a Smith-Fuzzy-PID control system for heat exchanger temperature control is designed. The Smith estimator compensates for the delay time, and the fuzzy PID control is used to speed up the response speed of the system, improve the control accuracy, and reduce the steady-state error. The controller simulation model is built in Simulink, and the simulation comparison is carried out with Fuzzy-PID, Smith-PID, and traditional PID control, and the interference test is carried out at the same time. The results show that Smith-Fuzzy-PID control has a fast response speed, no obvious overshoot of the system, strong robustness, and better improves the performance of the temperature control system to achieve the expected goal.

Multi-target search is a key part of UAV swarm cooperative reconnaissance. UAVs need to avoid threats in complex reconnaissance environments. In this paper, an optimization algorithm combining the artificial potential field with adaptive particle swarm algorithm (APF-APSO) is proposed for UAV swarm cooperative multi-target search mission. Firstly, we combine the PSO algorithm with the APF algorithm, adding a repulsive force field generated by the UAV's search path and using the force field function as a fitness function, which improves the obstacle avoidance and search efficiency of the UAV swarm. Secondly, a method for self-adaptive adjustment of personal cognition and social cognition factors is proposed, and an adaptive adjustment strategy of inertia weight is introduced to avoid the search falling into local optimum. The simulation results show that the proposed algorithm has higher search efficiency compared with other two PSO baseline algorithms in the same situation.

In the process of manual detection of Automobile Electrical Wiring Terminal(AEWT) defects in industrial production, there are problems such as low detection efficiency, time-consuming and labor-intensive. This paper puts an online detection algorithm forwards for the circularity of AEWT based on machine vision. Firstly, video images were obtained, and the key frame was extracted by the inter-frame difference method; Then, the adaptive Canny operator was used to extract the edges of images in different color channels fused with the AND operation; finally, the circularity of the AEWT is calculated employed by the edge image, and the defective products are judged according to the calculation result. Experiments show that the online detection algorithm that we proposed for the circularity of the AWET can effectively distinguish the defective products, the detection time is less than 30ms, and the accuracy is higher than 98%, which can meet the needs of real-time detection in practical application scenarios.

Coal is the most economical fossil energy source in the world, and the automatically drill pipe delivery of the drill rig can greatly reduce the labour intensity of workers. In this paper, we present a 2D vision sensor-based method for locating the delivery drill pipe on a horizontal directional drilling rig, which is achieved by the location of the drill arm position and identification of the drill pipe delivery position of a six-degree-of-freedom robot for underground drilling in coal mines. The identification of the drill pipe positions and delivery of it are realized by monocular vision based on the cooperative targets, which can effectively determine the robot posture positioning and precisely guide the robot to the drill pipe holder position on the drill rig. This method is successfully applied to the control of a practical industrial robot operator. The experimental results verify that our approach indeed enables the robot to automatically load and unload the drill pipe with excellent accuracy in practical applications.

Semiconductor packaging lead-frame has the characteristics of high density, high precision, refinement, and miniaturization. However, traditional manual detection has a series of problems such as difficulty, low efficiency, and high miss rate. Aiming at this industry problem, this paper develops a full-scale detection system and corresponding detection method of semiconductor packaging lead frame based on machine vision. The developed system is composed of a motion control platform and system, hardware control systems such as light source camera, and image algorithm platform. Through the optimized visual detection algorithm and the accurate correction method of workpiece attitude under motion measurement conditions, it can realize the one-click adjustment of several key dimensions such as frame length, width, loading area thickness, pin thickness, bending height, and aperture Micron level measurement, high precision, and fast speed, effectively ensure the genuine product rate and control the scrap rate, to assist semiconductor packaging enterprises to save labor costs and improve production efficiency.

A new design for a hardware system for photometric stereo-based robotic vision is proposed. In addition, a onefactor- at-a-time sensitivity analysis is performed to determine the optimal working distance for varying depths of field and feature depths for the photometric stereo (PS) sensor. The optimal working distance is defined as the distance at which the PS system is able to robustly achieve the best focus measure throughout the entire image. A cubic equation relating focus measure to working distance is found for each feature depth, and it is shown that good focus is achieved within the targeted range of working distances and feature depths. A validation study was also performed to show the results of PS using the designed sensor.

Mine waterjet cutting machine is mainly used for cutting steel strands in coal mines, and can also be used for cutting anchor rods and clearing rescue channels, etc. It has no cutting sparks, no thermal deformation, smooth and intact cuts, no pollution and sufficient abrasive sources, and is complementary to other cutting technologies. The system adopts premixed abrasive water-jet cutting technology, with the working pressure of 40 MPa, mass fraction of garnet of 20%, and target distance of 3~5 mm. With the case of Φ21.8 mm steel strand, its cutting speed is 45 mm/min and feedrate is 30 m/min. Under the assumed conditions, the pressure change model shows that, to ensure the cutting quality, steel strands should be cut only after the system pressure rises to close to the working pressure after about 100 s. After the system stops operation, its pressure needs to be reduced from the working pressure to the ambient pressure after 23 s.

The most commonly used implementation of handwritten digit recognition based on convolutional neural networks requires equipment with high computing power, which is not suitable for edge devices. Recently, spiking neural network (SNN) has received more attentions due to its low power consumption and real-time performance, but training SNN is very difficult. As a special SNN, liquid state machine (LSM) has the advantages of simple structure and easy training, so it is very suitable for handwritten digit recognition on edge devices. But it has no advantage in recognition accuracy. In order to improve the performance of LSM, its reservoir needs to be optimized. In this paper, an efficient local optimization strategy is proposed, improving the recognition accuracy of LSM by 11.8% with less the training time. In order to reduce the runtime, auto-encoder and feature screening are used to compress the input handwritten digit image. After feature compression, the input storage is reduced by 57%, and the runtime is reduced by 30%. This work provides an effective way to realize handwritten digit recognition on the edge devices.

Nickel-titanium (NiTi) shape memory alloy (SMA) has excellent application potential in aerospace because of its shape memory effect and super elasticity, but is still limited by poor machinability and weldability in traditional processing techniques. The coaxial powder feeding laser deposition technology opens a new window in the processing of NiTi SMA components. In this paper, prediction models between the process parameters (laser power, scanning speed and powder feeding rate) and process state parameter (melted pool temperature), deposition quality (track width, track height, microhardness) in NiTi alloy laser metal deposition based on the Back Propagation Neural Networks (BPNN) and Random Forest (RF) algorithms were estab-lished. Thirty single tracks were deposited and measured as training groups. The results show that the average prediction error based on the BPNN model for microhardness, track width, track height and melted pool temperature are 0.37%, 1.88%, 4.45% and 0.91%, respectively, which are better than the RF model. Then, BPNN model was further used to predict deposied quality under the combination of five new process parameters groups. Objective of this study was to provide a guidance for the subsequent optimization of process parameters for the improvement of the deposition accuracy of NiTi alloy parts.

Currently, there are advanced metal cutting processing methods such as electric discharge cutting, wire cutting, water jet cutting, but the metal cutting processing method is still very popular in the industry with its own advantages. The turning machining method is the same, in addition to the advantages, this method also has disadvantages. One of the disadvantages not achieving roughness as high as that of the finishing methods is the surface quality of the workpiece. One of the causes is the vibration of the workpiece and the tool during the machining process. This paper focuses on simulating the tool stiffness to the surface quality of the workpiece. The study has determined that the effect of the tool stiffness on the surface roughness is as a nonlinear function.

In this work, a compliant parallel tip-tilt motion stage design is proposed by employing novel bridge-lever composite mechanisms which can transmit the translational motion to the rotational motion. The variable axis flexure hinges are introduced in this particular design to achieve better translational-angular coupling, which improve the bearing capability and the motion range. Furthermore, a linear analytical model is also developed to describe the translational-angular behavior of the proposed mechanisms. Finally, finite element simulations are conducted to verify the analytical results and illustrate the dynamic characteristics of the tip-tilt stage. The results on the proposed design show significant potentials in applications such as active optics which require large range and high precision pointing, positioning and alignment of optical beams.

A novel quasi-zero stiffness air spring (QZS-AS) vibration isolation system of ship machinery is put forward in this paper. The system consists of the vertical air springs, the lateral air springs, the turning joints, the connectors and the mass. The vertical air springs are mounted vertically to bear the mass to provide positive stiffness. The turning joints, the connectors together with the lateral air springs provide the negative stiffness. The static analysis of the system is carried out and the optimal combination of the configurative parameters is derived. It shows that as the lateral air pressure increases or the connector length decreases, the negative stiffness provided by the lateral air springs increases to reduce the overall stiffness. Then, the dynamic equations are established. The stability analyses for the QZS-AS autonomous and non-autonomous system are carried out, and the stability conditions for both systems are given. It can be found that the novel system can realize quasi-zero characteristics to improve vibration isolation performance and maintain dynamic stability with proper parameters.

Difficult-to-machining materials are widely used in industry, due to their excellent material properties. However, the ultra-precision machining of difficult-to-machining is difficulty issue, due to their properties of super hard and super brittle. Ultrasonic assisted machining has a prominent role in processing difficult materials, because it combines the advantages of traditional processing and ultrasonic vibration. In this paper, ultrasonic assisted grinding processing method was used to machining silicon carbide material, using a self-made machining equipment, to explore the processing characteristics of ultrasonic assisted grinding in the processing of difficult-to-machining materials.

Strings are often used to increase the strength of the structures, such as stay cables in bridges. The string vibrations can affect the stability of the structure. In this paper, a method of a suppressing multi-order string resonance by controlling the length of the string to be continuously time-varying is proposed. The model of the time-varying length string is given. The vibration responses of the timevarying length string are solved by the finite difference method. The suppression effect of the time-varying string length method on the free vibrations and forced vibrations is analyzed. The results show that the controlling of the time-varying length of the string can effectively weaken the free vibration peaks and the forced vibration peaks of the string. It can suppress the multi-order resonance.

Asymmetric rolling is a process based on purposefully created differences in the circumferential speeds of the work rolls. For such a process, a degree of asymmetry is defined by a speed ratio (SR) of the work rolls. This research presents the experimental results of the effect of asymmetric rolling with high speed ratio (SR = 2.0; 2.5; 3.0) on the HSS AISI M2 behaviour during processing and on the change in the mechanical properties. Two types of asymmetric rolling were studied: 1) asymmetric warm rolling at temperatures 400, 500, 600, 700, 800 °C; 2) asymmetric hot rolling at 1100 °C. It was shown, that asymmetric rolling can be used as an effective technology for processing difficult-to-deform high-speed steels. Due to asymmetric rolling the number of passes was decreased by 5 times and at the same time the rolling force was also decreased by about 5 times compared with conventional (symmetric) rolling at the same temperature. The highest value of microhardness 790 HV (3 times higher than the initial value of microhardness) was achieved in HSS M2 after asymmetric hot rolling at 1100 °С and SR = 3. Carbides refinement with uniform distribution of them was also achieved. The results of investigation can be used for design of effective technology of manufacturing high-speed steels with improved microstructure and mechanical properties.

In the fault detection of mechanical rotating parts such as gears and bearings using data-driven methods, the working condition data sets detected by sensors have the characteristic of imbalanced proportion of positive and negative samples, which makes it difficult for machine learning classification methods to accurately identify fault samples. This paper proposes a method combining Synthetic Minority Over-Sampling Technique (SMOTE) with Generative Adversarial Networks (SMOTIFIED-GAN) to pre-process the training data. The input random noise of generator in GAN is replaced by the synthetic sample of SMOTE, which makes the samples more consistent with the real distribution, to solve the imbalanced class problem, thus improve the fault diagnosis ability of the classification model. This method is applied to experimental gearbox fault datasets. The classification ability and robustness of the detection model in different imbalance rates are analysed, and the results show that SMOTIFIED-GAN improved the detection rate. Compared with the traditional methods, the F1 score of fault detection results is improved, and the feasibility of the proposed method.

Aiming at the influence of the part manufacturing process energy consumption by the process sequence and the choice of the machine tool, the energy consumption modeling and optimization of part machining process based on feature was proposed. The concept of feature process element was proposed, incorporating feature combinations on the basis of comprehensive single feature description, and the corresponding sorting rules were established. Analyzed the machining capability and status information of machine tools, the model of the cutting energy consumption of the machine tool equipment and the energy consumption of the equipment empty cutting between the two features was established. Based on the correlation analysis of the direct and indirect energy consumption of the manufacturing process and the part feature process elements was carried out, the energy consumption correlation model of the part processing process was established, and the depth-first search algorithm was used to optimize the model. Taking a box part as an example to conduct application analysis, the feasibility and effectiveness of the above models and methods were verified

The monolithic multi-column load sensor is a leap of the traditional single-column load sensor in elastomer structure design, sensor fabrication technology and nonlinear compensation technology. This paper analyzes the advantages of the four-column sensor relative to the single-column sensor and the three-column sensor in detail, and makes a thorough theoretical analysis of the anti-deflection load capacity of a symmetrical four-column resistance strain sensor with integrated processing. By changing the pressure head of the sensor in different positions of the upper surface of the sensor, the anti-offset load performance of the four-column sensor was verified experimentally. At the same time, the test verification and simulation analysis were carried out on the force measuring platform composed of multiple fourcolumn sensors. The test data proved that the four-column sensor and the force measuring platform composed of it have good anti-offset load performance. It is very important for the application of column sensor

The innovation of high-speed cutting tools has promoted the rapid development of advanced manufacturing, especially intelligent equipment. At the same time, the development of intelligence has also had a certain impact on the development of high-efficiency CNC tools. With the continuous deterioration of working conditions, the material of the equipment is getting higher and higher. In order to solve this problem, the machine tool needs to be transformed. Therefore, there is a new development direction in design and manufacturing. Tool materials, coating technology and Geometry is an important factor in determining effective cutting results. From the point of view of digitization, this paper does a mathematical research on the manufacturing technology of CNC machine tools, and uses the powerful modeling ability of CAD software to carry out 3D modeling of the entire CNC tool and combine it with coating technology and cutting technology. The structural innovation of CNC tools not only greatly improves their processing capabilities, but also has further developed application fields.

With the continuous development of intelligent manufacturing technology, the machinery manufacturing industry has gradually entered the era of intelligence, which has transformed the traditional machinery manufacturing process and has become a trend in the current machinery manufacturing industry. The development of machinery manufacturing industry, in order to better apply the manufacturing technology of intelligent machinery, this paper summarizes the intelligent manufacturing technology, analyses the application of the current intelligent manufacturing technology in the manufacturing technology of machinery, and probes into the innovation path of the manufacturing technology of intelligent machinery, with a view to providing support for the innovative development of the intelligent machinery manufacturing process.

In order to make the breaking and closing operation of MCCB handle reach the optimal use state, this paper takes the MCCB with 4P pole number as the research object, focuses on the MCCB handle modeling, studies all problems in the MCCB handle operation with ergonomic methods, discusses the basic principles of ergonomics in depth, and integrates the national MCCB standard to study the contact relationship between the operation finger end and the handle during the force application process to analyze handle shape, size, edge chamfe, anti-slip area and panel depth design. Through measurement, test and systematic analysis, accurate research conclusions were drawn, improvement methods and suggestions. A series of improvements aim to enhanced the operability of the MCCB handle and improved the operating experience of the operator, which provides certain reference value for the research on improved handle modeling design of similar products.

Springback is a crucial factor in cold stamping that causes geometric inaccuracy of the stamped component after removal of tools. This study, for the first time, presents a novel application of a Convolutional Neural Network (CNN) based surrogate model to predict the thinning and springback behaviours for cold stamping. Datasets were created based on two cold stamping case studies, i.e., a U-bending case and an outer car door panel stamping case. The datasets were then applied to train the CNN-based surrogate models. The results show that the surrogate models can achieve near indistinguishable full-field predictions in real-time when compared with the FE simulation results. The application of CNN in efficient springback prediction can be adopted in industrial settings to aid both conceptual and final component designs for designers without having manufacturing knowledge.

Based on reverse engineering technology, this paper studies the design of automobile worm gear supercharger shell parts, uses a 3D scanner to obtain the three-dimensional coordinate values of automobile worm gear supercharger shell parts, processes and plans the collected point cloud data, completes the reverse modeling of the part surface through surface fitting and reconstruction, and then carries out innovative design and structural performance optimization according to the demand, and uses 3D printing technology to quickly form samples, Practice has proved that this innovative design method is feasible, and it also has guiding significance for solving the problems existing in the structure and performance of products and shortening the design cycle of products in the design of other products.