Structural Optimization Design Of Air-core Permanent Magnet Synchronous Linear Motors Based On Deep Neural Network Models

Air-core permanent magnet synchronous linear motor(PMSLM)can directly convert electrical energy into linear motion mechanical energy without an intermediate transmission structure.This PMSLM exhibits several advantages,such as no cogging effect,high dynamic response,and small size.Therefore,PMSLM demonstrates broad prospects in various fields,such as laser processing and additive manufacturing.However,due to the structural characteristics of linear motor itself,there are spatial harmonics and thrust ripple,which bring many challenges to the basic research and application development of linear motor.In high-precision CNC machine tools,the performance of permanent magnet synchronous linear motor determines the processing quality.In this paper,through the optimization design of PMSLM body structure,to ensure the high dynamic response and high-precision performance in motor operation.The main steps of structural design optimization of PMSLM include establishing a calculation model and implementing a global optimization.The calculation model reflects the nonlinear relationship between input structural parameters and output performance and requires fast and accurate calculation,thereby providing a fine basis for subsequent optimization.Therefore,the core problem of PMSLM structural design optimization is determining an efficient modeling method.In this paper,a regression modeling method of PMSLM based on the deep neural network(DNN)model is proposed,and the structural parameters of the motor are optimized by the immune clonal optimization algorithm under multiple working conditions to obtain the optimal structure parameter combination of PMSLM applied to laser cutting machine.The innovations and advantages of this study are summarized as follows: 1.This work is to increase the thrust while reducing the thrust ripple to achieve the best performance of PMSLM.2.The key structural parameters of the PMSLM are selected,which can provide an important basis for design optimization.3.A deep learning algorithm named DBN is introduced,which provides an idea for PMSLM analysis.The specific research is as follows:1.According to the topological structure characteristics of the two-sided air-core Permanent magnet synchronous line,the excitation effect of the permanent magnet is analyzed by the equivalent magnetization method,and the output performance analytical model of PMSLM is established,and the influence of the structural parameters on the motor performance is qualitatively analyzed.Then,the steps of establishing finite element model by finite element analysis method and parametric scanning technology based on finite element software are introduced.2.The sensitivity between the structural parameters and the output performance of PMSLM is analyzed,and the key structural parameters that have the greatest influence on the thrust and thrust fluctuation are selected.On this basis,the level factor table of the parameter variables is established,and the finite element model of PMSLM is established by the finite element analysis method.Within the limit of each parameter variable,the sample database of the relationship between the structural parameters and the output performance of PMSLM is obtained by the parametric scanning technology.3.In view of the shortcomings of support vector machine and other shallow neural networks,analytical modeling and finite element modeling in PMSLM modeling.By introducing deep learning algorithm,a fast calculation model of motor based on deep neural network is established,and the simulation results are compared with those of support vector machine and k-nearest neighbor algorithm to verify the superiority of DNN and provide a good foundation for subsequent iterative optimization calculation.4.Based on the DNN fast calculation model,a multi-objective optimization function is proposed to improve the thrust density and suppress the thrust ripple.Using the immune clone algorithm(ICA)to iteratively optimize the structural parameters of PMSLM,the optimal structural parameters of the motor under different working conditions are obtained.5.According to the optimal structural parameters,the PMSLM prototype is made and the prototype test platform is built.The output performance of PMSLM is tested under different conditions.The experimental results show that the method adopted in this paper can effectively improve the output performance of the motor,and further verify the accuracy and the advanced nature of the method.