To study the influence of pump motors under mixed operating conditions on the steering performance of tracked special vehicles,a performance evaluation model of pump motors under mixed conditions was established. The external characteristic equations for different leakage locations were solved,and key performance parameters such as volumetric efficiency and leakage flow were calculated.Combined with a steering dynamics model,the changes in pump motor performance under various conditions were analyzed. The model enabled the calculation of critical internal performance parameters of pump motors under complex conditions,providing a theoretical foundation for evaluating the steering performance of special vehicles. It is indicated by the result that steering performance was significantly affected by the degradation of the pump motor. As hydraulic system leakage flow increases and volumetric efficiency decreases,central steering time and turning radius increase.Under dynamic steering conditions,steering performance showed a nonlinear relationship with leakage flow and volumetric efficiency. The mixed-condition pump motor model effectively improved the accuracy and applicability of pump motor performance evaluation in special vehicle steering,providing a scientific basis for optimizing vehicle steering systems.

For the Green Distributed Heterogeneous Flexible Job Shop Scheduling Problem (GDHFJSP),a Cooperative Evolution algorithm based on Dueling-DQN (Dueling-DQNCE) is proposed. The objective is to minimize the maximum completion time and total energy consumption,selecting Pareto frontier solutions to achieve high-quality solutions. Initially,two initialization methods are introduced to effectively enhance the quality of the initial solution population.Subsequently,an activity scheduling scheme is utilized in the decoding stage to comprehensively explore the solution space and obtain high-quality solutions. Addressing the multi-objective problem,a rapid comparison method is proposed to efficiently obtain Pareto frontier solutions. Furthermore,ten knowledge-driven neighborhood search strategies are introduced,and Dueling-DQN is employed to intelligently learn and select appropriate local search strategies for each solution,accelerating population convergence. To validate the effectiveness of Dueling-DQNCE,it is compared with the state-of-the-art DQCE algorithm in 20 instances. The computational results demonstrate that Dueling-DQNCE outperforms Co-Evolution with Deep-Q-network (DQCE) in terms of computational resources and solution quality,confirming its effectiveness and superiority.

To address the multi-objective dynamic job shop scheduling problem and meet the real-time scheduling needs of manufacturing workshops in environments with variable scales, a method combining Proximal Policy Optimization (PPO) with GoogLeNet, named GLN-PPO, is proposed. This method constructs the state space of the scheduling problem using multidimensional matrices, designs an action space based on various priority rules, and devises a multi-objective reward function. To verify the effectiveness of the proposed algorithm, it is trained and tested in three environments: a static public environment based on common benchmark problems, a static real environment based on actual cases, and a dynamic real environment. Experimental results show that compared to genetic algorithms, GLN-PPO can provide high-quality scheduling results, meet the real-time scheduling requirements of enterprises, and adapt flexibly to environments with variable scales.

A 3D vision based robot polishing path planning method was proposed aimed at improving the accuracy and efficiency of robot polishing. Firstly,the structured light camera was used to obtain point cloud data on the surface of the workpiece. The RANSAC algorithm was used to remove the background,and the point cloud slicing algorithm was used to divide the row spacing. Subsequently,the Douglas-Peuker algorithm was used to dilute the path points,and the normal was calculated based on direct estimation method to generate accurate polishing path. In order to achieve stable force tracking,an adaptive variable impedance control strategy was designed. The performance of PID control method,constant impedance control method and adaptive variable impedance control method in planar,curved and complex surface environments was compared using the MATLAB Simulink simulation platform. The simulation results showed that the proposed adaptive variable impedance control method was superior to traditional methods in terms of force tracking accuracy and response speed. In the simulated polishing experiment,a polishing system based on FANUC M-10ia/12 robot and CX5020-0111 controller was built. The experiment verified the effectiveness of this method on surfaces of workpieces with different shapes,especially in complex curved environments.

The stiffness calculation of the bending anti-roll torsion bar device is one of the key points in the design of anti-roll devices,which is related to the arrangement space,installation method and structural form of the anti-roll torsion bar device. The calculation model of system roll angle stiffness is established what takes into account all key parameters,including bending radius,bending angle,torsion arm length,rubber joint stiffness,and support sleeve stiffness based on the principle of energy conservation. Based on the structure of bending anti-roll torsion bar device,the calculation model of roll angle stiffness is analyzed as an example,and the roll angle stiffness of the system is verified by combining with the bench test,and the test results show that the calculation model of the system roll angle stiffness is highly reliable and the error is only 3 %. Finally,the calculation model is used to explore the influencing factors of roll angle stiffness,such as key dimensional parameters and stiffness of rubber parts.

In order to improve the dynamic response characteristics and reliability of the automotive electromechanical braking system,the working principle of the electromechanical braking system was analyzed. Matlab/Simulink and AMESim software were used to establish the simulation model of magnetic field-oriented control and the physical model of mechanical actuator of permanent magnet synchronous motor,respectively,and a closed-loop control strategy based on expected clamping force to predict feedforward and fuzzy PID feedback was proposed. Based on AMESim and Matlab/Simulink software,a co-simulation platform was built,and step signals,square wave signals and trapezoidal signals were used to simulate the emergency braking,frequent braking and conventional braking conditions of the vehicle,respectively,and the effectiveness of the closed-loop control strategy of clamping force was verified by simulation and comparative analysis.The simulation results show that compared with the traditional method,the proposed strategy significantly improves the dynamic response speed and control accuracy of the system.

In order to solve the current problem that most studies focus on the development of related hardware,communication protocol and upper computer software,too reliance on manual calibration parameters,the lack of automatic optimization calibration of the research problems, first by using an orthogonal trial design,the control parameter with great influence on the test index is selected as the input factor for the subsequent calibration. Secondly,a combined optimization algorithm is constructed to optimize the key parameters,with the combination of hardware and software,such as MicroAutoBox,Isight and MATLAB/Simulink,the automatic optimization calibration platform is designed. Finally,through the bench test verification,the automatic optimization calibration bench of new energy vehicles is built, the calibration time is reduced by 74.86 % compared with the manual calibration,realizing the high efficiency and automation of the calibration process. Under the target working condition,the power consumption of 100 kilometers is reduced by 4.3 % compared with that before the automatic optimization calibration.

The inaccurate selection of the suspension position of the exhaust system can easily lead to abnormal noise and vibration during the driving of the car,and it is extremely important to accurately select the position of the suspension point of the exhaust system. The position value extracted from the vibration mode of the exhaust system is weighted and processed by the Root Mean Square (RMS) method to obtain the root mean square value method suspension point layout of the vibration mode displacement. The position of the suspension points arranged by the Average Driving Degree Of Freedom Displacement (ADDOFD) method and the suspension points obtained by the root mean square method are compared and verified. The results show that the root mean square method of suspension point arrangement is effective in SUV. The root mean square method has a higher distribution efficiency than the average driving degree of freedom displacement method.Shortened design cycles.

In order to improve the driving stability of distributed drive electric vehicles under various steering conditions and coordinate the rotational speed difference control between wheels, a differential steering control strategy integrating both a composite rotational speed control method and a slip rate control method is proposed. Firstly, the relationship between the optimal slip rate and the road adhesion coefficient is investigated, and their correlation is fitted to ensure the system operates near the optimal slip rate. Then, a fuzzy controller is employed to process the inputs-including the deviation between the actual yaw rate and the desired yaw rate, the deviation between the actual side slip angle and the desired side slip angle, and the vehicle speed, and output a coordination coefficient for the two control methods. Under stable conditions, the rotational speed control method is utilized to achieve desired wheel speed tracking via sliding mode variable structure control. Under unstable conditions, the slip rate control method calculates the additional yaw moment by sliding mode variable structure control, while a single/dual-wheel switching controller allocates the torque. Finally, simulation results shown that, under high-speed and low-adhesion conditions, the proposed control strategy reduces the maximum yaw rate and side slip angle by 37.96 % and 52.82 % respectively, compared to a single control method, significantly improving vehicle handling stability.

The generalized growth and remanufacturing of waste mechanical products is a multi-level and multi-particle remanufacturing process from waste mechanical products to performance improvement,function expansion,or appropriate reuse. There are many kinds of feasible product sets for generalized growth and remanufacturing of waste mechanical products,the product information is complicated and the information connection is poor,which affects the retrieval and selection efficiency and accuracy of waste mechanical products and their parts remanufacturing target products. To solve the problem,a modeling method of waste mechanical product information based on fusion semantic information was proposed. Considering the correlation between parts in waste mechanical products and the material,structure,and failure information of the parts themselves,integrating semantic information and topological graph theory,the product information that affects the selection of generalized growth and remanufacturing can be accurately described and integrated into the product information graph. Through the case of parts retrieval,the validity of the proposed product information model was verified by comparing it with the existing information model.

In order to solve the problem of low load capacity and static rigidity of a single-hole throttled aerostatic bearing,three cross-sectional forms of throttling grooves,namely triangle,rectangle,and trapezoid,were designed to form a composite throttling method. The gas thrust bearing was taken as the research object to build the gas film characteristic model under the small hole throttle and compound restrictors with different cross-sectional forms of throttling groove. The FLUENT software was used to carry out simulation calculations,to get the bearing capacity,static stiffness and outlet flow rate of the gas thrust bearing.On this basis,the influence of parameters such as the cross-sectional form,cross-sectional area,and gas supply pressure of the throttle groove on the static characteristic of the composite throttling gas thrust bearing was analyzed. The results show that the throttle groove can effectively increase the bearing capacity,static stiffness and outlet flow rate;different cross-sectional form of throttle groove has a different impact on the static characteristic of the bearing,low film thickness trapezoidal cross-section throttle groove shows better bearing capacity,static stiffness and outlet flow rate;within a certain range,as the cross-sectional area increases,the bearing capacity,static stiffness and outlet flow rate show a tendency to increase first and then decrease; the gas supply pressure has a significant impact on the static characteristics of the throttle groove composite throttling gas thrust bearing,as the gas supply pressure increases,the bearing capacity,static stiffness and outlet flow rate continue to increase,but the magnitude gradually of the bearing capacity and static stiffness slows down.

With the rapid advancements in machine vision technology,intelligent welding techniques have witnessed increasingly widespread applications in the field of mechanical manufacturing. Three-dimensional models were constructed by using structured light cameras. An autonomous robotic welding method eliminating manual teaching procedures was developed,demonstrating automatic generation of high-precision welding paths on complex geometric profiles of cast nail rollers.Initially,the point cloud data obtained from 3D reconstruction underwent segmentation and preprocessing through algorithms such as surface fitting,pass-through filtering,region growing,and edge extraction. Subsequently,a mathematical model was established for the edge point cloud and centroid of the ceramic column to determine the welding points. Finally,the Principal Component Analysis (PCA) algorithm was employed to optimize the welding path in 3D space,enabling autonomous planning of the entire welding trajectory for the curved workpiece. Sample data indicates that the overall system error is less than 0.5 mm. Engineering test results demonstrate that the weld appearance exhibits no significant defects,fulfilling the autonomous welding production requirements for cast nail rollers through the Metal Active Gas arc welding (MAG) process. Valuable guidance and reference for the autonomous surfacing welding engineering applications of cast nail rollers with curved surfaces was provided.

To address the issue of occlusion-induced model uncertainty during the assembly of industrial robotic arms,which complicates the accurate estimation of hole part poses,an improved method combining the YOLOv8 model with adaptive unscented kalman filtering,named SHUK-SVSF algorithm,was proposed. Firstly,to reduce response loss caused by occlusion and improve hole part pose detection accuracy,the SEAM module was introduced into the output layer of YOLOv8 model′s neck. Secondly,a design based on the Sage-Husa algorithm adaptive unscented kalman filter was proposed,which adjusts the measurement noise covariance matrix in real-time,thereby enhancing the stability and accuracy of pose estimation in dynamic environments. Finally,the introduction of variable structure gain within the unscented kalman filtering leverages the precision of unscented kalman filtering and the stability of smooth variable structure,further improving pose estimation performance. Experimental results demonstrate that the modified YOLOv8 method exhibits high detection accuracy in occluded scenarios,while the proposed SHUK-SVSF algorithm meets the real-time requirements for visual guidance in robotic arm assembly processes and achieves high accuracy and robustness in pose estimation under model uncertainty conditions.

In order to determine the optimal light source position parameters of visual inspection system for high reflective metal processing size,processing accuracy and surface quality,an optical system optimization design model based on non-imaging optics theory is proposed. According to the principle of illumination uniformity and Taguchi experimental design method,the optical system of high reflective metal surface visual inspection is optimized by adjusting the light source position parameters ( the irradiation angle of the light source,the inclination angle of the light source and the distance of the light source). The results show that under the optimal light source position parameters,the uniformity and standard deviation of the illuminance distribution in the field of view imaging plane are 88.89 % and 0.036 2,respectively. Combined with the imaging experimental results,it can be seen that the illuminance uniformity of the high reflective metal surface under the optimal light source position parameters is better,which is helpful to the high-quality imaging of the visual inspection of the high reflective metal surface,and provides technical support for the improvement of the visual inspection accuracy of the high reflective metal surface.

Aiming at the problems of large differences in diamond blade defect features and limited defect samples,an anomaly detection algorithm for diamond blade based on the improved U-Transformer feature reconstruction model was proposed.This method only needs to be trained with normal samples to complete the detection and localization of abnormal areas. Firstly,a frozen pre-trained deep Convolutional Neural Networks (CNN) model was used to extract multi-scale fusion features,amplifying the differences between normal and abnormal images. Then,the encoder-decoder feature reconstruction model based on the U-Transformer structure was constructed,and the feature similarity between the reconstructed features and the input features was calculated to generate the similarity response map. Finally,to eliminate the noise responses in the normal region,the similarity response map was corrected to obtain the anomaly score map by estimating the anomaly scale factor using the Multi-Layer Perceptron (MLP) network. The experimental results demonstrate that the proposed method achieves an Image AUROC index of 0.989 and a Pixel AUROC index of 0.992 on the diamond blade defect dataset,meeting the requirements for diamond blade anomaly detection.

Fault diagnosis of rolling bearings is related to the safety and stability of equipment operation.Traditional convolutional neural network is used to diagnosis malfunctions,whose model arithmetic is too large and prone to overfitting phenomenon.It caused some problems such as poor diagnostic accuracy,end-to-end modeling with poor trustworthiness and so on. In view of this,it was supposed to propose a rolling bearing fault diagnosis method based on improved Symmetrized Dot Pattern (SDP) and FasterNet-GCAM network. Firstly,the one-dimensional vibration signals were processed by wavelet thresholding noise reduction,and then fed into the SDP method optimized by the Pearson′s image correlation coefficient method to generate the SDP images. FasterNet network by adding the idea of partial convolution was constructed into an improved SDP-FasterNet model for further feature extraction to complete the classification and diagnosis of different faults in the bearing. Lastly,in order to verify the credibility of the model′s decision-making in the process of image recognition,Gradient-weighted Class Activation Mapping (Grad-CAM) was combined with the FasterNet network,which can highlight the important parts of the SDP graph which are relevant to decision making. The experimental results show that the proposed method has faster convergence speed and stronger robustness than other methods. And the results′ diagnostic recognition accuracy had achieved 99.20 %. It improved the interpretability and credibility in the diagnostic process,which provides a diagnostic model with good feasibility and robustness in the field of fault diagnosis.

Spherical mobile robots,owing to their advantages such as strong stability,high flexibility,and low energy consumption,have significant applications in unstructured scenarios like rescue search and terrain exploration. Currently,domestic and foreign scholars have conducted relatively few studies on the deformation structures of spherical mobile robots. Their deformation forms and applicable scopes are categorized and summarized from 4 aspects in the study:deformation based on key structures,deformation based on topological theory structures,deformation of amphibious spherical mobile robots,and deformation of jumping spherical mobile robots. The enhancement of maneuverability and terrain adaptability in spherical mobile robots is acknowledged to hold referential significance.

With the deepening of the development plan of intelligent equipment in important fields such as national defense and medical care in China,the high precision,high flexibility,complex motion ability,and high human-machine integration ability of robots have become urgent needs in reality. The active and passive flexible adaptive ability of their joints has become the core technology of research. In response to this issue,taking the flexible implementation form of joints as the main line, systematically reviewed the research status of flexible joint variable stiffness technology for robots at home and abroad are from five aspects:using elastomers to achieve flexibility,viscoelastic materials to achieve flexibility,magnetic materials to achieve flexibility,driving technology to achieve flexibility,and hybrid technology to achieve flexibility.The principles,typical design structures,and characteristics of various types of flexible joint flexibility technology are summarized,providing reference and inspiration for further innovation of robot flexible joint technology.