Aiming at the problems of unstable attitude of mountain unmanned operation chassis under complex roads and poor adaptability and robustness of traditional control methods,a chassis attitude control strategy based on Newton-Raphson-Based Optimizer (NRBO) algorithm,eXtreme Gradient Boosting (XGBoost) algorithm and Twin Delayed Deep Deterministic policy gradient (TD3) algorithm was proposed. Firstly,the seven-degree-of-freedom active suspension vibration model environment was built;then the state prediction model of NRBO-XGBoost was trained,the state prediction model was added to the TD3 algorithm and the attention mechanism was added to the network to enhance the decision-making ability and adaptive ability of the TD3 intelligences in complex environments,and at the same time,the reward function was designed and the TD3 intelligences were trained to realize the chassis attitude control in complex road environments; finally,simulations were carried out based on Matlab 2023a/Simulink software. The simulation results show that the chassis attitude strategy method based on the improved TD3 can effectively suppress the attitude change of unmanned operation chassis under complex roads,and the pitch angle,lateral inclination angle,and vertical displacement are suppressed by 61.4 %,84.9 %,and 84.9 %,respectively,which significantly improves the smoothness;compared with the traditional DDPG,PPO,and TD3 reinforcement learning control strategies,with the improved TD3 algorithm the pitch angle is improved by 49.1 %,7.4 % and 37.2 %,respectively,the lateral inclination angle is improved by 83.3 %,36.5 % and 34.7 %,respectively,the vertical displacement is improved by 70.7 %,77.5 % and 64.0 %,respectively,and the vertical displacement acceleration is improved by 67.7 %,42.1 % and 49.7 %,respectively,which provides a better control effect with better adaptability and robustness.

Based on the special configuration and use requirement of aero-engine blisk,a new high-performance and low-cost silicone rubber-based fluid abrasive was developed. The processing effect of silicone rubber-based fluid abrasive on the surface of blisk was investigated. The effects of main abrasive parameters (abrasive mesh,abrasive mass ratio,abrasive material) on viscosity and processing performance were analyzed. A single factor experiment was used to analysis the variation of the surface roughness,material removal rate,and surface microstructure of the blisk under different abrasive parameters. The optimal abrasive parameters were obtained. The abrasive contamination detection and verification experiment of the blisk workpiece were carried out. The results show that the new fluid abrasive has good rheological properties,processing performance,and high efficiency. Using 80^# silicon carbide abrasive,and the mass ratio of silicone rubber matrix to abrasive is 1∶1.4,better processing quality can be obtained,and the reduction rate of surface roughness is more than 50 % after 54 cycles. The self-made new fluid abrasive can effectively improve the surface integrity of the blisk parts. It will not pollute the surface of leaves. The accuracy of the leaves will not be destroyed,which can meet the requirements of the blisk.

Aiming at the problem of long cycle and high cost of complex product design, a module division method considering part reuse is proposed. Firstly, the three correlation attributes of function, physics and reusability of complex product parts are analyzed, and a comprehensive correlation matrix between parts is constructed. Combined with complex network theory, it is mapped into the relationship network model between parts. Then, an improved Louvain community discovery algorithm is proposed, which solves the problem of too many small communities by selecting seed nodes and realizes the division of complex product modules. Finally, the chip removal machine is taken as an example to divide the module, and compared with the division scheme obtained by Louvain community discovery algorithm and spectral clustering algorithm, the effectiveness and feasibility of the proposed method are verified.

To effectively reduce the complexity of CNC milling process planning for complex surfaces and achieve its automatic generation,a recommendation model for CNC milling process planning based on the random forest algorithm was proposed. First,surface features and process information of CNC operations from existing machining instances were extracted,and a CNC process knowledge base was constructed by integrating CNC machining theory. Then,a recommendation model for CNC milling process planning based on the random forest algorithm was designed,which takes the extracted surface features of the part to be machined and the machining requirements as inputs,and outputs the CNC machining planning parameters with the highest score,thereby generating an appropriate CNC milling process planning recommendation scheme. The validity of this model was verified using corresponding CAD/CAM software.

In view of the lack of functional types of the current logistics warehousing management system,low human-computer interaction,information transmission is not timely,the overall situation of the warehouse is difficult to control,low ease of operation,it is difficult to carry out real-time control and intervention of the status of the logistics warehousing system. Based on this,from the perspective of digital twin model construction,it constructs the digital twin model of intelligent logistics and warehousing management from four levels,constructs the three-dimensional model of warehousing components through software such as SolidWorks and 3Ds MAX,and carries out the motion logic modeling of transportation equipment in warehousing. By using Unity 3D visualization software and C# programming language,the digital twin logistics warehousing visualization system was developed,and the system realized the expected functions of logistics warehousing visualization,data visualization,scene roaming,collision warning and so on. It is expected that relevant work can provide suggestions for the transformation of intelligent logistics and warehousing industry to digital intelligence,and provide references for the construction of digital twin model standard system in related fields and industries.

The application of deep reinforcement learning algorithms in palletizing robotic arm trajectory planning suffer from slow learning rate and poor robustness. To address the above problems,a Twin Delayed Deep Deterministic policy gradient (TD3) algorithm based on improved Azimuthal reward function (A) is proposed for trajectory planning of robotic arm. First,the mathematical model of the palletizing robot is established in Cartesian coordinate system and its kinematic analysis is carried out. Second,for the problems of slow learning rate and poor robustness,based on the relative directions and positions of the robotic arm and the obstacles,an improved Azimuthal reward function combined with Twin Delayed Deep Deterministic policy gradient (A-TD3) algorithm is designed for the palletizing robotic arm trajectory planning,which enhances the robotic arm target oriented search,and improves the learning efficiency and robustness. Simulation results show that compared with the TD3 algorithm,the average convergence speed of A-TD3 algorithm is improved by 11.84 %,the average reward value is improved by 4.64 %,the average extreme deviation is decreased by 10.30 %,and the trajectory planning time is lower than that of the mainstream RRT and GA algorithms,which verifies the effectiveness of the A-TD3 algorithm in the application of palletizing robotic arm trajectory planning.

Aiming at the demand for stable walking of quadruped robots,a stability analysis and planning research of the backward gait of quadruped robot was carried out. Firstly,the structure of the legged robot was introduced. The motion coordinate system was established by D-H method,and the auxiliary action parameters were preliminarily determined. Then,the stability margin index was used to characterize the static and dynamic stability of the quadruped robot,and the relationship between the center position of gravity and time was calculated according to motion parameters and joint drive functions. Next,in the gait design process of quadruped robots,the influence of leg and head to tail movements on the center of gravity of the quadruped robot was analyzed,and the full-cycle action sequence of the backward gait was planned through active and auxiliary actions,and the action parameters of the backward gait were determined. Finally,the feasibility and effectiveness of the proposed method were verified by virtual prototype simulation and physical prototype experiment,which provides theoretical method support for gait analysis of quadruped robot.

In order to improve the efficiency and reliability of industrial robots,the topology optimization design of a certain type of industrial robot′s large arm was conducted. The fuzzy satisfaction variable weight coefficient method combined with the membership function was used to assign the weights coefficient for the multi-objective topology optimization of the large arm,which can remove the human factor in the traditional multi-objective topology optimization. A finite element model of the large arm was established and its correctness was verified by strain measurement experiments and acceleration measurement experiments,dynamics simulation of industrial robots was conducted by Adams software to obtain the force on the big arm under three working conditions,and the single-objective topology optimization design of static stiffness under three working conditions and the first three inherent frequency was performed,then the fuzzy multi-objective topology optimization structure of the large arm was obtained. Based on optimization results the large arm was reconstructed,the stress of the traditional optimized arm increases instead of decreasing under working condition 2,and the magnitude of other sub-objective optimization is not as good as that of the fuzzy multi-objective topology optimization. The large arm optimized with fuzzy multi-objective topology reduces the weight by 12.2 % compared with the original arm,while the stiffness under the three working conditions and the first and the third order natural frequency are all improved to different degrees,proving the optimization method′s effectiveness.

Flexible grippers,with their ability to conform to various shapes,offer significant advantages over rigid grippers,particularly when handling irregularly shaped or delicate objects. However,traditional flexible grippers often struggle with limited load capacity due to material and structural constraints,making them less effective for certain applications. A novel design for a variable stiffness flexible mechanical finger was introduced,based on the principle of granular body obstruction. This design incorporates a granular interference layer,which significantly enhances the stiffness and load capacity of the finger through active particle interaction.Unlike conventional pneumatic flexible grippers,this design allows for precise control of the finger′s bending by adjusting air pressure,optimizing its fit around irregular objects. Simultaneously,the stiffness of the finger can be fine-tuned by controlling the vacuum pressure within the granular interference layer. As the vacuum pressure increases,the walls of the compartment compress the particles more tightly,resulting in greater stiffness and enhanced load-bearing capacity.This active particle interference approach offers a key advantage over passive systems, the ability to adjust the finger′s stiffness independently of its bending angle. Moreover,as the bending angle increases,the range of stiffness adjustment also expands. Both theoretical analysis and experimental results demonstrate that this variable stiffness flexible finger,utilizing vacuum-controlled granular obstruction,effectively allows for independent control of bending angle and stiffness,enabling the flexible gripper to perform a wider array of robotic grasping tasks.

A sliding mode fault-tolerant control strategy based on extended state observer was designed to solve the problem of actuator failure,considering external interference and actuator input saturation. Firstly,the extended state observer was used to estimate the complex interference including the fault. The extended state observer is independent of all the state measurements and mathematical models,and can accurately estimate the external interference and the actuator fault. Aiming at the input saturation problem of actuator,a nonlinear anti saturation compensator was designed to compensate the input saturation overflow effect in real time. Then a finite-time controller based on non-singular terminal sliding mode was proposed to realize the precise control of the manipulator system. The convergence time of the sliding mode variable was bounded and independent with the initial state of the system. Based on Lyapunov stability theory,the finite-time stability of closed-loop system was proved. Finally,the superiority of the proposed controller was verified by simulation and comparison with the control method combining the extended state observer and boundary layer technology in the literature.

To solve the problems of using traditional A^* algorithm for AGV path planning,such as multiple search nodes,redundant and uneven paths,and inability to handle random obstacles in complex environments,a real-time path planning and obstacle avoidance method for Automated Guided Vehicles (AGVs) was designed by integrating improved A^* algorithm and Dynamic Window Approach (DWA) algonthm. Firstly,an adaptive evaluation function was designed based on the traditional A^* algorithm evaluation function,allowing the algorithm to adaptively adjust according to the surrounding environment during the search process,thereby improving the speed and flexibility of the algorithm. Secondly,in response to the problem of redundant nodes in the path,the Floyd algorithm was used for bidirectional smoothness path optimization,deleting redundant nodes to retain key turning points,reducing the number of turns,effectively improving path smoothness,and fully ensuring the stability of AGV operation. Finally,the improved A^* algorithm was combined with the DWA algorithm to achieve effective integration of global optimization and dynamic obstacle avoidance in path planning. This comprehensive method not only enhances the path planning ability of AGV in complex environments,but also improves obstacle avoidance performance,providing a more reliable solution for the practical application of AGV.

Addressing the problems of the traditional A^* algorithm in complex environments,including poor searching effect,tortuous paths,too close corners with obstacles and poor obstacle avoidance,a robot path planning algorithm integrating improved bidirectional A^* and Dynamic Window Approach (DWA) was proposed. Firstly,the bidirectional search mechanism and adaptive heuristic function were introduced to improve the search speed of the algorithm.Secondly,the line segment intersection method and safe distance determination were used to eliminate redundant points to ensure a safer and smoother path. Then,azimuth and target distance determination were added to DWA to enhance the obstacle avoidance ability of the algorithm. Finally,DWA was used in real-time obstacle avoidance in segments of the optimized global path.Simulation results show that the fusion algorithm can generate smoother paths within the safety boundary in complex scenes,effectively avoid obstacles and quickly reach the target point.

With the increasing global number of vehicles, end-of-life cars pose growing environmental and resource challenges. Disassembly, recycling, and remanufacturing are key solutions. However, their complex structure creates sequence dependence among disassembly tasks, while part deterioration introduces uncertainty, affecting efficiency. To develop a practical disassembly approach, it first formulates a stochastic programming model to represent uncertain information, establishing a multi-objective sequence-dependent disassembly line balancing problem under uncertainty, optimising time, energy consumption, and idle time equilibrium. Next, an improved bees algorithm with innovative search operators is proposed, incorporating a roulette wheel strategy for adaptive weight updates. Finally, a case study on a crank train from a scrapped engine validates the model and algorithm.

In order to study the influence of bellows parameters such as the angle,fillet radius,and straight edge length of bellows on the wall thickness distribution of TRB pipe non axial feeding hydraulic forming V-shaped bellows, finite element simulation is used forming experimental for corrugated pipes with different parameters. The uniformity of wall thickness is used as the evaluation index,a response surface prediction model is established between bellows parameters and positive,negative,and total differences. Aiming at the problem of not meeting the uniform distribution of wall thickness when forming large angle bellows under theoretical design conditions,a method of changing is proposed for improvement. The results indicate that the angle has a significant impact on the distribution of wall thickness,with an increase in the angle leading to a decrease in wall thickness uniformity,the fillet radius and straight edge length on wall thickness distribution is not significant impact. When the total difference is less than 20.0 %,the maximum angle value is 32.1 degrees,and the total difference is less than 10.0 %,the maximum angle value is 25.2 degrees under theoretical design conditions. Taking a certain parameter combination as an example,after improvement,the total difference decreased from 28.4 % to 15.2 %,forming V-shaped corrugated pipes with larger angles and uniform wall thickness is feasible.

Titanium alloys,due to their high specific strength,excellent heat resistance,and corrosion resistance,are widely used as primary materials in blisks for the aerospace industry. However,the complex geometric shapes,stringent machining precision requirements,and the unique characteristics of titanium alloys present significant challenges during the milling process.These challenges include tool wear,surface quality of the machined parts,and machining efficiency. First,the geometric model of the titanium alloy blisk was established and CNC programming was performed based on the machining features of the blades and flow paths. Subsequently,the milling process for titanium alloy blisks was analyzed,and appropriate tools and machining parameters were selected for experimental validation. Finally,the profiles of the machined blades were measured and analyzed,aiming to provide theoretical guidance and technical support for the efficient and high-quality machining of titanium alloy blisks.

Electromagnetic clutch is an important part in the automobile production process. Aiming at the problems of small defect size of its adsorption surface,complex background texture and existing algorithms can not achieve the diversity of defects,a lightweight target detection algorithm based on improved YOLOv8n was proposed. EMA attention and partial convolution were integrated in the backbone network,and a lightweight C2F-PE module was designed to improve the C2F structure and to enhance the feature extraction ability of the network. In order to promote richer feature fusion between the same scales,the Attention-based Intra-scale Feature Interaction (AIFI) module was introduced to replace the SPPF layer to capture more fine-grained information.A small object detection layer was added to the neck network,which effectively fused the shallow feature information and improved the model perception of small objects. The Slim-neck module was introduced to improve the neck network,which lightened the model while maintaining the detection accuracy of the network. The experimental results showed that compared with the YOLOv8n algorithm,the improved algorithm achieved an mAP@0.5 of 94.6 %,which was an increase of 4.5 %. The number of parameters was reduced by 13.3 %,and the detection speed reached 81 f/s. The algorithm effectively balanced detection accuracy and speed,meeting the needs for real-time detection in electromagnetic clutch production.

To solve the problem of difficult recognition of rolling bearing failure modes due to the complex and variable operating conditions of power plant equipment,a rolling bearing fault acoustic pattern recognition method based on IPNCC-MCSKNet was proposed to realize the efficient recognition of rolling bearing faults under the operating conditions of variable rotational speed. Firstly,the acquired bearing acoustic signals were preprocessed,noise reduced,and feature difference integrated to form Improved Power-Normalized Cepstral Coefficients (IPNCC). Then the various voiceprint features including IPNCC were extracted to construct multi-channel input features,and the mechanism that the Selective Kernel (SK) convolution module can adaptively adjust the size of the convolution kernel was utilized to establish a multi-channel selective kernel network model (MCSKNet). Finally,voiceprint pattern modeling and fault recognition were carried out on different fault forms of rolling bearing samples. The experimental results showed that the proposed model achieved an average diagnostic accuracy of 95.99 % in the diagnostic task under multiple variable speed conditions,which was 13.98 % to 26.55 % higher than other deep learning models,and the model was more robust. The results can provide new ideas for rolling bearing acoustic feature extraction and fault diagnosis.

To address the ride comfort degradation caused by hub motor rotor eccentricity in Hub Motor Driven Electric Vehicles (HMD-EV),an H₂/H_∞ Robust Model Predictive Control (RMPC) strategy for an active suspension system based on a Particle Swarm Optimization algorithm (PSO) is proposed. Firstly,a vertical-lateral dynamic coupling model of the HMD-EV is established,and the key factors such as radial and tangential magnetic field distortions and Unbalanced Electromagnetic Forces (UEF) resulting from the rotor′s static eccentricity are thoroughly analyzed. The dynamic interaction between the rotor eccentricity in the permanent magnet synchronous hub motor and the electromechanical-magnetic coupling under external disturbances is studied. An active suspension PSO-H₂/H_∞ hybrid controller,optimized by feedback gain matrix (K),is designed. Simulation results demonstrate that the controller effectively mitigates the adverse effects of electromechanical-magnetic coupling in HMD-EV,significantly improving the vehicle′s ride comfort and smoothness.

In order to further improve the sinusoidal nature of the air-gap magnetism waveform of eccentric magnetic pole Permanent Magnet Synchronous Motor (PMSM) and enhance the efficiency of motor optimization design, an optimization design method based on analytical calculation and GA is proposed for the eccentric pole of PMSM. Firstly, considering the effects of tooth-tip saturation, stator slotting, pole-to-pole leakage and edge effect and magnetic circuit saturation on the air-gap flux density, the analytical calculation model of air-gap flux density under parallel and radial magnetization of eccentric poles to satisfy the subsequent optimization calculations is deduced, respectively. Then, the pole structure parameters of this model are optimized by GA. Finally, the accuracy of analytical calculation and optimization results are compared and verified by finite element simulation. The results show that the analytical calculation model obtained under this method has high accuracy, the sinusoidal nature of the air-gap magnetic density waveforms of the motor optimized for both parallel and radial magnetizing poles is improved, the cogging torque is reduced, and the efficiency of the optimized design is improved.