In order to achieve high performance dry cutting of difficult-to-machine materials,the DLC/TiAlN composite coating tool with both antifriction and wear resistance was prepared by using TiAlN with high hardness and high bonding strength as the support layer and DLC coating with self-lubricating properties as the top layer. At the DLC/TiAlN interface, liquid-phase assisted laser texturing was performed to enhance the interlayer bonding strength.Using dry cutting 316 stainless steel test, 4 kinds of cutting tools,including TiAlN coated (NC) tool,untreated DLC/TiAlN composite coating (UC) tool,air laser treated DLC/TiAlN composite coating (AC) tool and liquid phase assisted laser treated DLC/TiAlN composite coating (LC) tool,were used to study cutting properties such as force/heat,chip morphology,coating peeling,tool bonding etc, when dry cutting stainless steel under different cutting parameters. The test results showed that the cutting forces of UC,AC and LC tools were reduced in different degrees compared with NC tools.When the cutting speed is 200 m/min,the total cutting force of LC tool was the lowest,which was reduced by 26.5 %. At the same time,the front and back tool face wear of LC tool was the least,and the coating integrity was the best.

The residual tensile stress generated during the processing operations of laser welding in 45Mn sheet has adverse effects on their strength,toughness,and fatigue life.Various ultrasonic impact processes were employed to enhance the surface of the sheet and a multi-objective optimization method for the ultrasonic impact process parameters of sheet was proposed. Firstly,a dataset of surface residual stress under different impact process parameters was obtained through finite element impact simulation.Then,based on the simulation dataset,a nonlinear mapping relationship between impact process parameters and surface residual stress was successfully established using the IDBO-BP neural network. Comparing the IDBO-BP neural network with BP,GA-BP,PSO-BP,and DBO-BP neural networks,it was found that the IDBO-BP neural network achieves higher accuracy in predicting surface residual stress of the sheet,with MAE and R² evaluation metrics reaching 0.068 3 and 0.997 4,respectively,indicating the effectiveness of the model in predicting residual stress after ultrasonic impact. Finally,considering the ultrasonic impact process parameters as design variables and aiming for minimal residual stress,minimal impact current,and minimal impact time,a Pareto optimal solution set of residual stress,impact current,and impact time corresponding to the ultrasonic impact process parameters was obtained by combining the IDBO-BP neural network and the PSO algorithm. The results demonstrate that the optimized impact process effectively improves processing efficiency and processing energy efficiency.

The job-shop scheduling problem is a classic NP-hard combinatorial optimization problem,and the quality of scheduling directly impacts the operational efficiency of manufacturing systems.In order to obtain a better scheduling strategy with the goal of minimizing the maximum completion time,a Deep Reinforcement Learning (DRL) scheduling method based on Proximal Policy Optimization (PPO) and Convolutional Neural Network (CNN) is proposed. A three-channel state representation method is designed,with 16 heuristic scheduling rules selected as the action space,and the reward function is equivalent to minimizing the total idle time of machines. In order to enable the trained scheduling strategy to handle scheduling instances of different scales,Spatial Pyramid Pooling (SPP) is applied in the convolutional neural network to convert feature matrices of different dimensions into fixed-length feature vectors.Computational experiments are conducted on 42 Job-Shop Scheduling Problem (JSSP) instances from the public OR-Library. The results of the simulation experiments show that the proposed algorithm outperforms single heuristic scheduling rules and genetic algorithms,achieving better results than existing deep reinforcement learning algorithms in most instances,and with the smallest average completion time.

In a discrete workshop production environment,the degradation process of equipment is influenced by external random shocks,and production planning is affected by equipment conditions. Production planning can be jointly optimized with maintenance scheduling. To address this,a production scheduling optimization model that considers equipment degradation was proposed. To solve this model,a degradation model considering random external shocks was used,with the goal of minimizing the maximum completion time. A Tri-Population Evolutionary Genetic Algorithm (TPEGA) was designed for this purpose. This algorithm employs three distinct populations with different functions to collaboratively search for the optimal solution,restricting the population size while maintaining diversity among high-quality solutions. To avoid falling into local optima,a local optima probability model and an individual discarding strategy were devised. Additionally,to optimize the initial population,a greedy selection strategy considering the equipment with the highest load and process time was proposed. Experimental results demonstrate the effectiveness of the algorithm and the feasibility of the model.

Intelligent manufacturing is a common trend in the development of the world′s manufacturing industry,and the digital twin,as an important enabling technology to promote the implementation of intelligent manufacturing,is a key link to intelligent manufacturing. To deal with the problems of uncertain events and low transparency of production information,a flexible job-shop scheduling optimization method based on digital twin simulation was proposed. Firstly,the digital twin shop scheduling framework was designed,and the AnyLogic software flexible job-shop digital twin simulation model was constructed. Secondly,considering the maximum completion time,energy consumption and total load of equipment,a flexible job-shop scheduling model was established,and an improved NSGA-Ⅱ algorithm was proposed to solve the problem.Multi-strategy mixed population initialization method was adopted,and different cross-mutation strategies were adopted for process and machine coding. Finally,the effectiveness of the proposed method was verified based on the standard calculation examples and a manufacturing example of engine cylinder head.

In order to solve the problem that the state prediction is difficult to perform due to the complex production process and frequent abnormal disturbances in the discrete workshop,a collaborative state prediction method of workshop production-logistics based on spatio-temporal feature was proposed. Firstly,based on the production-logistics operation logic and real-time manufacturing data,the production-logistics collaboration relationship was analyzed,and the prediction index of the production-logistics collaboration state was determined. Secondly,according to the spatio-temporal feature relationship between production and logistics,a production-logistics time sequence graph model was established,and then the spatio-temporal fusion network based on Graph Attention network-Gated Recurrent Unit (GAT-GRU) was used to predict the prediction indicators of the cooperative state. Finally,a typical mixed-flow production workshop was used as a case study,and the experimental results show that the proposed prediction method was better than the deep neural network,denoising autoencoder,gated recurrent unit and other models in terms of accuracy and efficiency,and can more effectively realize the production-logistics collaborative state prediction.

In order to improve the practicality of the staircase cleaning robot,a staircase cleaning robot with scalable rotating robotic arm was designed,which used the rotation of the robotic arm to go downstairs,and at the same time,a combination of a telescopic rod and a tensioning device was used to achieve the scalable function,improving its adaptability to stairs of different heights. The staircase cleaning robot mainly consists of cleaning structures,scalable downstairs structures,control systems,and other parts. By designing corresponding control strategies for the robot,it can achieve discrimination of stair edge signals,control of downstairs,single-layer stair surface obstacle avoidance,stair end obstacle avoidance,stair kick surface height detection and other functions.The above control strategies were executed alternately to achieve the robot's cleaning of the entire stair surface. The scalable experiment of the robotic arm shows that the robotic arm can extend and shorten to different lengths (24~29 cm). The experiment of going downstairs with different heights shows that the robot can adapt to stairs with different heights between 10~20 cm,and the success rate of going downstairs is 100 %. The experiment of stair kick surface height detection shows that the robot can adjust the appropriate length of the robotic arm based on the height of the stair kick surface.

Currently,in the shipbuilding industry,robotic arm welding operations have gradually replaced traditional manual labor.To improve the working efficiency and stability of robotic arms,a time-optimal trajectory planning method for robotic arms based on the Terminal Crossover and Steering-based Particle Swarm Optimization (TCSPSO) algorithm was proposed. First,a 5-7-5 polynomial interpolation function was constructed to fit the motion trajectory in the joint space of the robotic arm,with the goal of optimizing the robotic arm′s motion time,and a constrained optimization model was established. Then,the Augmented Lagrange multiplier Method was applied to convert the constrained optimization problem into an unconstrained one. To avoid local optima,the TCSPSO algorithm was used to solve the problem. Finally,simulation experiments were conducted in MATLAB software,resulting in the optimal motion time and smooth trajectory of the robotic arm. The results indicate that this method can effectively shorten the motion time of the robotic arm while ensuring its stability during movement.

Scroll compressors have broad prospects for development in the field of miniaturization,reducing the size of scroll plate is an important optimization goal of miniature scroll compressor,and selecting a scroll profile with a compact structure is crucial. Using a variable base circle radius involute,instead of the traditional fixed base circle radius involute,can reduce the diameter and tooth height of the scroll plate,resulting in a reduction in the radial and axial dimensions of the scroll plate.In order to study the influence of the size reduction in two directions on the mechanical properties of the scroll plate,fluid-heat-solid multi-field coupling analysis on fixed base circle radius involute scroll plate and variable base circle radius involute scroll plates with reduced diameter and tooth height was performed using ANSYS software. The mechanical properties of three scroll plate were compared using stress and deformation as indicators. The research results indicate that reducing the tooth height by using variable base circle radius involutes can significantly improve the mechanical performance of the scroll plate. Using variable base circle radius involute to reduce the diameter can better meet the needs of lightweight and miniaturization,but the mechanical performance will decrease.

The rolling dielectric elastomer made of silicone rubber has been widely used in soft robot field because of its high frequency resonance. The high conductivity and very low stiffness of liquid metal can effectively guarantee the resonance response of dielectric elastomer.However,the current Rolling Dielectric Elastomer Actuator (RDEA) prepared with liquid metals cannot achieve the resonance phenomenon. It presents a method to obtain a flexible electrode with oxide and liquid metal co-existing by stirring liquid metal and scraping coating. Then,the static and dynamic properties of the Spring Rolling Dielectric Elastomer Actuator (SRDEA) prepared by the process and the traditional carbon-based electrode are compared,and the reliability and low stiffness characteristics are demonstrated. Finally,based on this process and structure,a bristle crawling robot was made,which can achieve a crawling speed of more than 1 times the body length per second.

In order to overcome the problems of large off-axis amount,large vector height difference and low surface accuracy faced by traditional machining technology when machining the off-axis parabolic mirror with equal thickness,a machining method based on slow tool servo technology was proposed,which is applicable to large off-axis amount,large vector height difference and high machining accuracy. The method adopted the coordinate translation strategy,which effectively solved the problem of large off-axis amount existing in the off-axis paraboloid; a new angular coordinate rotation method was used to ensure that the vector height difference at the far and near axis ends of the off-axis paraboloid is close to 0,so as to achieve the goal of equal thickness. In addition,the method of equal angle-equal arc length was used to discretize the contact points between the tool and the machined surface,and the tool radius compensation in Z direction was carried out for these contact points to generate the machining path,making the machining accuracy higher. RSA6061 aluminum alloy off-axis parabolic mirror parts (diameter of 25.4 mm,radius of curvature of 50.8 mm,off-axis angle of 90°) were processed by the processing method proposed. An equal thickness off-axis parabolic aluminum alloy mirror with surface accuracy of RMS of 0.124λ(λ= 632.8 nm) and surface roughness of Sa of 1.952 nm has been successfully fabricated. The machining method not only makes the machining process more concise,but also helps to improve the machining quality of the workpiece.

To solve the problem of difficult to accurately and quickly locate the protecting cover seal screw in the scene of automatic installation for single-phase smart meters,a rotating detection algorithm GROOVE-YOLO based on improved YOLOv8obb was proposed,which could obtain central position and rotating angle of the target screw through a single detection. Firstly,the P2 detection head was added and the P5 feature layer was deleted to build a tiny object detection network,which boosted small object detection capability while simplifying the network structure; secondly,the bidirectional feature pyramid network was fused and learnable weights were applied to improve the effect of multi-scale feature fusion and enrich the model feature expression; finally,the global attention mechanism was employed to highlight the key information in features and further enhance the detection accuracy. The experimental results showed that precision,recall,average precision of the improved algorithm reached 93.2 %,90.5 %,95.7 %,which were increased by 13.5 %,17.1 %,14.3 % compared with the original model. The average pixel distance error of the center point was 1.4 pixels,the average rotating angle error was 4.8°,and the detection speed reached 119 f/s. The proposed method can effectively improve the position accuracy of the protecting cover seal screw and meet the actual installation requirements.

In order to meet the requirements of deep subpixel localization and strong noise resistance for image edge detection in machine vision precise measurement,a Canny-Devernay subpixel image edge detection algorithm based on bicubic interpolation was proposed. Firstly,the median filtering algorithm was used to reduce the noise in the acquired images. Secondly,the image edges were subdivided through bicubic interpolation. Finally,the contour generation step was optimized and Ostu algorithm was applied to obtain an adaptive threshold,which was passed into Canny-Devernay algorithm to achieve accurately extracting subpixel edges of image. An experimental system was developed by high-resolution industrial camera and high-performance computer,and experiments were carried out on the OpenCV image and the acquired images of USB interface plug-in parts as samples,which was done with camera calibration accuracy of 0.009 8 mm/pixels.The experimental results show that the average edge detection error of the proposed algorithm is 0.006 85 mm,which is less than 0.7 pixels,and the calculation time deviation is 7.68 ‰. The edge positioning accuracy,noise resistance and compute stability of the proposed algorithm were better than the Canny algorithm,the algorithm based on Zernike moments and the Canny-Devernay algorithm. The proposed algorithm can better meet the requirements of the stability,reliability and high-precision image edge detection for machine vision precise measurement,and would be used to develop a new type of machine vision dimensional precise measurement device.

During the manufacturing process of aluminum profiles,defects such as scratches and dirt spots may occur on the surface due to factors like materials or processing techniques,directly affecting the usability of aluminum profiles. It analyzes the characteristics of surface defects in aluminum profiles and compares existing deep learning-based object detection algorithms. Based on the YOLOv8 network model,an aluminum profile surface defect detection algorithm integrating GhostBottleneck and attention mechanism is proposed. By introducing Ghost into the Bottleneck layer and replacing some of the convolutional structures in the backbone network with DWConv,the complexity of the model is reduced while ensuring detection accuracy. Furthermore,the ECA attention mechanism is added to the YOLOv8 detection head module to enhance the detection accuracy of the model. It conducts experimental verification,and the experimental results show that the accuracy of the improved algorithm reaches 0.932,representing a 5.9 % improvement compared to the basic YOLOv8 algorithm. Moreover,the number of model operation parameters is reduced by 24 %. The overall performance meets the industrial requirements for the accuracy and speed of defect detection in aluminum profiles.

To improve the roll stability and ride comfort of mining dump trucks under harsh road conditions,a connected oil-gas suspension system was selected as the research object. An integrated model of the connected oil-gas suspension system,road surface,and vehicle body for mining dump trucks was established using the AMESim and Simulink joint simulation platform. By analyzing the impact of suspension system structural parameters on stiffness and damping,5 optimization design variables were selected. To enhance the diversity of the initial and subsequent generations,a rapid Nondominated Sorting Genetic Algorithm-Ⅱ (NSGA-Ⅱ) based on orthogonal experimental design was proposed and applied to the multi-objective optimization of the connected oil-gas suspension system.Simulation results indicate that the improved NSGA-Ⅱ algorithm provides better Pareto solutions,with roll stability and ride comfort of the mining dump truck improved by 26.69 % and 24.83 %,respectively,under D-class random road input.

The prediction of degradation trends in rotating machinery rolling bearings faces issues such as reliance on prior knowledge and low identification accuracy. To address these issues,a method that integrated Principal Component Analysis (PCA) with Random Forest (RF) and Gated Recurrent Unit (GRU) network for rolling bearing degradation trend prediction was proposed. Firstly,a set of high-dimensional features based on multivariate statistics was optimally selected. Dimensionality reduction and clustering were conducted by PCA to construct health indicators of the rolling bearing. Secondly,the health indicators were constructed to serve as basis,the RF model was introduced to fit the rolling bearings degradation curve. Finally,a PCA-RF-coordinated GRU network model of the rolling bearing degradation trend prediction was established to complete the rolling bearings status assessment. It is verified from experiment that the health indicators of the proposed method can effectively reflecting the degradation status of the rolling bearing,with the time trend of 0.999 1. Furthermore,it is shown that the PCA-RF-coordinated GRU model can accurately predict the degradation trends of the rolling bearing. The maximum root mean square errors for single-step and multi-step predictions on different datasets are 0.018 4 and 0.047 8,respectively.

Aiming at the problem of decreased control accuracy caused by uncertainties such as nonlinear mechanical friction,external disturbances,and model errors in the dual-axis cradle turntable system,an improved sliding mode control law was designed using an extended state observer. Firstly,the spatial coordinate system relationship among the base,cradle and turntable of the dual-axis cradle turntable was analyzed,and the dynamic models of the cradle and turntable were established. Then,the uncertainty in the system was extended to the system state,and the designed extended state observer was used to estimate the uncertainty. Finally,the improved sliding mode control law was designed to dynamically compensate for the influence of uncertainty,effectively suppressing the phenomenon of chattering and achieving high-precision control of the dual-axis cradle turntable. The simulation results show that the designed extended state observer can accurately estimate the uncertainty in the dual-axis cradle turntable,and the maximum estimation error is only 0.01 (°)/s². The maximum tracking errors of the improved sliding mode control law for turntable rotation angular velocity and cradle rotation angular velocity are only 0.02 (°)/s and 0.03 (°)/s,respectively. The test results show that the proposed improved sliding mode control method has stronger robustness and higher accuracy compared to adaptive iterative learning control method and command filtered adaptive backstepping control method,and the maximum positioning errors for turntable rotation angle and cradle rotation angle are only 0.04° and 0.05°,respectively,greatly improving the control accuracy of the dual-axis cradle turntable.

It conducts multi-step optimization design of the weak structural member column based on the dynamic performance analysis of the entire machine tool. The column′s external structure and internal ribs are optimized using finite element method for the single column. The results show that the original rib pattern and distribution of the column have unreasonable design structures. Therefore,the column is topologically optimized,and the column's external structure and internal rib patterns are redesigned based on the data obtained from the dynamic performance analysis and the material distribution after topological optimization. The rib patterns and sizes are also analyzed for their parameter correlation,and a response surface optimization model is established. The Multi-Objective Genetic Algorithm (MOGA) (a variant of NSGA-Ⅱ) is used to optimize the key dimensions for multiple objectives,and the optimized entire machine is compared with the model to confirm the comes about. The comes about appear that the first-order common recurrence of the optimized column has expanded by 10.2 %;the maximum resonance peak in the X direction has decreased by 15 %,in the Y direction by 25 %,and in the Z direction by 54 %. Furthermore,the wave peaks in the X and Y directions have been shifted back to some extent. Therefore,the optimized machine tool has improved its static and dynamic performance.