Fir tree blade root milling cutter is an important tool for processing blade roots and hubs in the steam turbine industry.Due to its complex cross-sectional shape,the traditional machining technology has the shortcomings of complex process,low efficiency and difficult to guarantee accuracy. In order to solve the process problems above,a process strategy for roughing of five-axis turning and milling compound machining center and finishing of five-axis tool grinder was proposed. Firstly,taking a fir tree-shaped blade root milling cutter as an example,the structural characteristics of the five-axis turning and milling compound machining center were determined,and the process route of the five-axis turning and milling compound machining center roughing and the finishing of the five-axis tool grinder was determined,and the process technical documents were compiled respectively. Secondly,ESPRIT TNG turning and milling composite programming software was used to compile the roughing toolpath trajectory,and NUMROTO software was used to compile the finishing toolpath trajectory. Finally,the toolpath trajectory was converted into numerical control code for simulation processing,and the simulated numerical control code and grinding code were imported into the five-axis turning and milling compound machining center and the five-axis tool grinder to complete the practical processing. By testing the tool profile,angle and appearance of the finished machined tool,all technical indicators meet the design requirements,and the processing efficiency increased about 24.4 %,which verified the effectiveness of the process scheme.

The chip evacuation performance of the drill directly affects the cutting performance.In order to address the issues of chip entanglement and vibrations encountered during the cutting process of indexable insert drills,and to prevent resulting processing abnormalities and poor hole wall quality,a combination of theoretical analysis and simulation was utilized to conduct an in-depth study on the impact of the cross-sectional shape and helix angle structure of the chips flute on the drill body rigidity.Experimental validation was conducted to demonstrate the excellent rigidity and chip evacuation performance of the designed drill.Simulation experimental results analysis showed that when the core diameter is consistent,the maximum stress on the drill body is 761 MPa,the strain is 0.003 367,the drill body head lateral displacement is 0.397 1 mm,and the rotation change is 0.006 500 rad;when the variable helix angle is between 20°-16°,the maximum stress on the drill body is 695 MPa,the strain is 0.002 688,the drill body head lateral displacement is 0.364 1 mm,and the rotation change is 0.006 140 rad,achieving a good balance between drill body rigidity and chip evacuation performance.The analysis indicates that as the helix angle increases,the difference in leaf width increases,and the core diameter decreases,the drill body rigidity becomes poorer. The designed drill demonstrated better chip evacuation performance without chip entanglement under experimental conditions,and exhibited high hole diameter consistency and good hole wall quality.

Aiming at the problems of laser vision robots during the welding process,the laser stripe segmentation accuracy of the weld seam was reduced due to noise interference and the inaccurate welding,an improved U-Net weld seam feature recognition method was proposed. The improved U-Net algorithm incorporates Mobile-Net as the underlying network,thereby augmenting the network′s capability for feature recognition and reducing the number of model parameters. Efficient channel attention network was added between encoding and decoding to achieve weighted fusion of features.The model adopted a mixed loss function to balance the proportion of laser stripes in the image. The network model was implemented on the experimental platform for seam tracking in welding robots. Experimental results showed that the Mean Intersection over Union (MIoU) of the improved U-Net algorithm was 89.83 %,the Pixel Accuracy (PA) was 99.54 %,the Mean Pixel Accuracy (MPA) was 97.28 %,and the image processing time was 0.209 s.Compared with other algorithms,it has better segmentation accuracy and faster processing speed.It can be more effectively applied to robot welding scenarios with interference.

In order to adapt the production demands of hybrid flow shop, a two-stage dynamic scheduling method based on machine learning is proposed.In the offline mining phase, based on historical data, the improved random forest algorithm is used to establish a knowledge mapping network from the production state of the manufacturing system to the optimal scheduling rules, which can be mined for online decision-making, thus skipping the warm-up phase to improve the scheduling efficiency and optimize the scheduling scheme. In the online scheduling phase, the reinforcement learning algorithm is used to analyze and train the real-time data of the flow shop state. And it optimizes the strategy selection according to the system state′s dynamic changes of the system state to achieve the adaptive and fast response capability to the perturbation events. The simulation experiment verified that the two-phase dynamic scheduling approach combining data mining and reinforcement learning is more feasible and effective for fully utilizing manufacturing data and scheduling the manufacturing execution process online.

Flexible stretchable electronics are widely used in fields such as wearable electronics and soft robots,etc.In order to achieve a high degree of integration,electronics between different layers,etc.,need to be connected together by vertical interconnections,and thus it is crucial to achieve stable vertical interconnections. A method that can directly print low-melting-point liquid metal columns by electrohydrodynamic jetting was proposed,which can realize vertical interconnections of flexible circuits. By controlling the printing parameters,different heights of metal risers can be printed to fabricate vertical intercon-nections of different heights. In order to test the stability and reliability of the vertical interconnection structure,the fabricated circuits with vertical interconnection channelsion were subjected to tensile and bending tests,respectively,which showed that the circuits had a stable resistance response under hundreds of bending and tensile cycles. A wireless induction coil with LED lights based on a multilayer structure was fabricated to demonstrate the capability of electro hydrodynamic jet printing for vertical interconnections.

Fabrication of conformal sensors for curved surfaces is important in structural health monitoring of three-dimensional components. However,attaching sensors to a curved surface by transfer printing is difficult to ensure the accuracy of position and orientation. Therefore, a five-axis motion platform was combined with the Direct Ink Writing (DIW) process to directly printing conformal sensor arrays on non-expanded curved surfaces. Firstly, a sensor printing process and developed motion control algorithm was proposed. Secondly,a multilayer structure of sensor array was printed on a UAV blade to detect the strain,and the data was compared with a finite element model. The results showed that the monitoring data matches well with the simulated data in terms of trend and error is less than 15 %,proving that the sensor is capable of monitoring the strain on the surface of the components. Finally,the sensor array was printed on the surface of a propeller with greater curvature to illustrate the high adaptability and the wide range of application scenarios of the scheme.

In order to solve the problems of long research and development cycles and low digitization in Exhaust Gas Recirculation (EGR) module development,a method for constructing a digital twin system for the EGR module was proposed. Firstly,accordingto the theoretical framework of digital twin model construction,an EGR module digital twin framework was designed from five aspects—physical model,digital model,virtual-to-real connection,state mapping,and twin services. Secondly,the construction of the digital twin model was achieved through three steps—solid three-dimensional modeling,front end scene building,and physical script mounting. Considering the characteristics of multi-source heterogeneity of twin data,a method for data acquisition and transmission was proposed,and visualization mapping was implemented.Finally,using the Unity 3D virtual engine as a platform,the above functions were integrated,and the EGR module digital twin system was developed. Experimental validation demonstrates the feasibility and effectiveness of the system,providing new insights for the digitized assistant design of EGR modules.

In order to reduce the time and computing power requirement of the collaborative planning process of multi-welding robots,a collaborative planning method for multi-welding robots based on finite state machine theory was proposed. Taking the automobile side wall assembly as the welding object, the hybrid Genetic Algorithm-Gravitational Search Algorithm (GA-GSA) was used to solve the welding joint allocation problem of each welding robot in the working cell and complete the welding path planning. The collision zone between robots during the welding process was identified to accurately establish the mutually exclusive zone,and Stateflow tool was used to model and simulate the multi-welding robots system.According to the actual working conditions, the priority input of each welding robot in the system and the time required for each welding joint to complete the welding task can be set to obtain the optimal adjustment scheme of welding robot action in the system under this working condition. Two simulation cases verified the feasibility and effectiveness of the proposed method,and the results show that the proposed method can quickly and effectively complete the collaborative planning of multi-welding robots.

In response to the issues of parameter perturbations,external disturbances,and filtering errors affecting trajectory tracking accuracy in single-joint flexible robotic arm control,a command-filtered sliding mode control method with error compensation was proposed. Firstly,the dynamic equation of the single-joint flexible robotic arm with uncertainties was presented and a backstepping sliding mode control was designed. Then,a second-order command filter was introduced to obtain the derivative of the virtual control signal,thereby avoiding the ″computational explosion″ issue associated with traditional backstepping control. Finally,an error compensation mechanism based on Lyapunov theory was used to eliminate filtering errors generated by the command filter. Simulation results demonstrate that when the system experiences parameter perturbations and external disturbances,the proposed error compensation command-filtered sliding mode control achieves trajectory tracking accuracy improvements of 19.8 % and 74 % compared to command-filtered sliding mode control without error compensation and error compensation command-filtered control methods,respectively.Filtering errors are reduced by 44.4 % and 31.4 %,respectively.

In order to solve the problem that traditional underdriven manipulators can only passively adapt to the morphology of grasping objects,a rope and Shape Memory Alloy (SMA) spring-based differential drive manipulator was proposed,which can actively control the grasping morphology of the fingers according to the characteristics of the objects to be grasped,and is more in line with the grasping characteristics of the human hand. The kinematic model of the finger was developed using the D-H method and the workspace of the finger was analyzed;the static mechanical model of the manipulator was established according to the principle of moment equilibrium,and the influence laws of the driving force of the drive rope and the equivalent stiffness of the finger joints on the grasping pattern were elucidated,and the results of the theoretical analysis were verified by ADAMS simulation;finally,the prototype was tested experimentally,and the research results show that according to the contour size of the object,the typical grasping actions such as enveloping and pinching can be accomplished by actively controlling the driving force of the drive rope and the equivalent stiffness of the finger joints,so as to realize the stable grasping of objects of different shapes and sizes.

Path tracking control is a key technology of port Automated Guided Vehicle (AGV),and the existing path tracking control methods have problems such as difficulty in maintaining high accuracy on reference paths with large curvature mutations such as U-turns,and poor real-time performance. To this end,a path tracking control method for port AGVs based on preview LMPC is proposed based on Linear Model Predictive Control (LMPC) without preview,where the linear expansion point is changed to the preview point in front of the AGV. The path tracking control method for port AGVs based on preview LMPC is compared with Nonlinear Model Predictive Control (NMPC) and LMPC without preview through the joint simulation.The simulation results show that the accuracy of the preview LMPC is high,the displacement error is no more than 0.060 2 m,which is comparable to that of the NMPC and significantly better than that of the LMPC without preview,and the reduction of this index can reach 93.25 % under the same working condition. The real-time performance of the preview LMPC is also good,and the solution time in each control cycle is no more than 6.020 3 ms,which is comparable to the LMPC without preview and significantly better than the NMPC,and the reduction of this index can be up to 72.52 % under the same working condition.

Aiming at the problems of long path,large number of iterations and local optimization of the basic dung beetle optimization algorithm in path planning,an improved dung beetle optimization algorithm was proposed. Firstly,Chebyshev chaos mapping was used to optimize its initial population in order to improve individual quality. Secondly,the golden sinusoidal algorithm was introduced into the position update strategy of foraging dung beetle,which can improve the convergence speed of the algorithm and improve the effectiveness of the algorithm search. Finally,dynamic weights were added to the stealing dung beetle position to expand the search range of the algorithm and reduce the probability of the algorithm convergence in advance.Using the raster map method to establish the environment map,simulating the working environment of the handling robot,using MATLAB simulation software,the improved dung beetle optimization algorithm and the basic dung beetle optimization algorithm were simulated in different raster maps,and the improved dung beetle optimization algorithm in three different raster maps was obtained by analyzing the simulation data,which was 31.25 %,10.31 % and 8.59 % shorter than the basic dung beetle optimization algorithm in terms of average path length,respectively. The average number of iterations decreased by 22.26 %,12.19 % and 26.24 %. The average number of inflection points decreased by 35.11 %,9.52 % and 35.09 %.It is proved that the improved dung beetle optimization algorithm has better effect in the path planning of the handling robot,and the feasibility of the improved dung beetle optimization algorithm is verified.

To enhance the regenerative braking energy recovery rate of pure electric vehicles while ensuring effective braking performance,a refined approach was proposed to optimize the fuzzy control strategy using an improved whale algorithm. Introducing battery State of Charge (SOC),vehicle speed,and braking intensity as fuzzy control inputs,with regenerative braking proportion coefficient K as the output,the control parameters were optimized using an improved whale algorithm,thereby enhancing the proportion of braking power provided by the front axle motor. Concurrently,refining the adaptive weights of the whale algorithm prevents the algorithm from converging to local optima during the iterative process. Through simulation analysis,the enhanced fuzzy control strategy proposed has been validated to not only improve energy recovery compared to both the pre-optimized and traditional control strategies but also ensure braking effectiveness under NEDC conditions.

Comparative experiments were conducted on the electric discharge wire cutting machining of cylinder type parts through conventional and layered liquid supply.Research found that the processing of electrode wire in the forward direction through layered liquid supply can be achieved in the entire process of spark discharge,the problem of insufficient internal liquid supply inside cylinder type parts by reverse electrode wire transport was solved,the interpolar working fluid flow rate and flow rate were increased,the chip removal capacity was enhanced,the discharge gap was enlarged and the interpolar discharge conditions was improved.The processing of electrode wire in the forward direction raised the spark discharge efficiency and stability and play an important role in the improvement of workpiece cutting speed and surface processing quality.

In order to improve the static and dynamic characteristics of vertical machining center,the weak link of machine tool was identified by modal analysis and harmonic response analysis,and it was found that the unreasonable design of the gantry frame structure leads to large deformation of the machine tool in the vertical direction.Aiming at the unreasonable design of the gantry frame,topological optimization and two-dimensional section optimization were used to determine the material distribution;based on MOGA genetic algorithm,the mass,maximum static deformation and first-order natural frequency of the gantry frame were taken as optimization objectives,and the key dimensions of gantry frame were optimized by multi-objective parameters;then,the optimization results were compared with the prototype. The results show that the maximum static deformation of the gantry is reduced by 34.61 %,the mass is reduced by 7.64 %,and the first-order natural frequency is increased by 6.54 %;at the same time,the maximum amplitudes of the whole machine in the X,Y,and Z direction are reduced,and the position of each wave peak has different degrees of backward,the static and dynamic comprehensive performance of the machining center are improved;the optimization effect of gantry frame is verified.

In order to carry out efficient part changes to adapt to the changing design requirements for serialized part models with similar structural features, the idea based on feature decomposition is proposed to realize its reconstruction. The model decomposition is realized by refining the composition of Boolean segmentation ring, which resolves the model into a Boolean combination of feature bodies; establishing the model weighted attribute adjacency graph, segmenting out the feature subgraphs, and obtaining the geometric topological relationship between the segmentation block and the model as a whole, so as to characterize the connection between the segmented elements of the model; by judging the priority of the segmentation ring segmentation and by deducing from the decomposition process in reverse direction, the face-shell obtained by segmentation is restored into a feature body, and the model is established. The directed graph of design history characterizes the modeling process; the parametric reconstruction of the model is realized based on geometric constraints and positioning constraints. It is verified that the method can realize the rapid change of serialized part models when the part models are reconstructed under the premise of geometric and positioning constraints.

In the process of precision boring,the deep hole structure parts of the bottle cavity,represented by the aerospace engine vortex shaft,are prone to chip removal difficulties and low processing efficiency. The formation of chips during boring was studied,and the influence of different process parameters on chip morphology and chip removal effect was mastered. Based on the Box-Behnken design test,the finite element analysis method was used to simulate the deep hole machining process of the bottle cavity,and the chip curl radius prediction model was established with the process parameters as the independent variables,and the reasonable process parameters were selected to optimize the chip removal. The accuracy of the chip curl radius prediction model was verified by actual machining test. The test results show that the similarity between the test value and the predicted value was more than 90 %,which proves the validity and accuracy of the chip curl radius prediction model.

Superalloys have good high temperature resistance,oxidation resistance and corrosion resistance,which are widely used in aerospace,energy,and other fields. However,due to their high tensile strength,low thermal conductivity,large resistance to plastic deformation during cutting,serious cutting heat agglomeration and work hardening,superalloys are considered difficult-to-machine materials,so it is challenging to make small holes on superalloy by cutting. The small hole machining of GH605 cobalt-based superalloy was carried out by means of helical milling.The orthogonal experiment of micro helical milling was designed and carried out to explore the influence of key milling parameters (spindle speed n,feed speed v_f,and pitch s) on cutting force F,cutting temperature T and surface roughness Ra of hole wall. Based on the range analysis results of the experimental data,the appropriate cutting parameters were selected,and the micro helical milling of straight holes and inclined holes were successfully carried out.

It introduces the working principle,system structure,fitting interpolation algorithm and simulation analysis of a NC cutting device for intersecting line curves,designs a NC cutting device for intersecting line K-groove with X、Y、Z、U、W and 5-axis linkage,controls the running trajectory and cutting angle of the cutting torch through the five-axis linkage interpolation algorithm,and establishes a mathematical model for controlling the cutting trajectory of intersecting line K-groove by combining Newton iteration method and the principle of spatial analytic geometry.

In the traditional optical angular displacement measurement method,there are some problems such as complex optical path and weak anti-interference ability. Therefore,a novel angular displacement measurement method based on optical coding imaging is proposed. Firstly,the method uses the principle of converting gray code into binary for reference,evenly distributes two circles of RGB (Red,Green,Blue) blocks on the circular surface with a diameter of 50 mm,and converts the color information around the outer circle target point into binary number corresponding to the absolute position by detecting the color information. Then,in order to reduce the noise,Gaussian filtering and median filtering are used to preprocess the image,and the Hough circle algorithm is used to locate and track the target point to obtain its starting and ending positions,so as to calculate the coarse angle. Finally,the coarse angle is compensated by the angle between the position of the target point in the last frame of the video and the final position to achieve angle measurement. By building the experimental platform and making the sensor,the accuracy measurement and stability experiment of the experimental device are carried out in the 360° measurement range. The resolution is 16 bits,the measured mean square deviation is 0.23°.The new angular displacement measurement method proposed brings new ideas for the field of angular displacement measurement.

Based on the principle of maximum tangential force and Fourier function,it takes Darrieus vertical axis wind turbine of type H as the research object. In other words,the optimal angle of attack is determined based on the principle of maximum tangential force,and the theoretical curve of pitch variation is generated. Meanwhile,in order to ensure the periodicity and ease of implementation of the rule of pitch variation,the control points are discretized from the theoretical pitch curve, and Fourier function is used for quadratic optimization and fitting. Then,the influence of variable pitch rule on the aerodynamic performance of vertical axis wind turbines is studied from the aspects of static start-up performance and dynamic output performance by numerical simulation. The results show that the power coefficient of vertical axis wind turbines is significantly improved at different tip speed ratios,and the improvement effect is particularly significant when the tip speed ratio is low,and the peak power coefficient is increased by 13.5 % at the optimal tip speed ratio of 1.25. At the same time,the static starting performance of vertical wind turbine under each phase angle is greatly improved,and the maximum static torque coefficient is increased by 261.1 % compared with that before variable pitch. The research results can provide a reference for the optimization of aerodynamic performance of vertical wind turbines.