Optimal Design Of Switched Reluctance Motor Based On High Performance Surrogate-assisted DE Algorithm

With the development of industrial society,the performance of electric equipment is required to be higher and higher,so the research and optimal design of novel electric machine is very important.However,due to the disadvantage of switched reluctance motor such as large torque ripple,obvious noise and so on,its application and generalization is hindered.At present,domestic and foreign researchers focus on improving the shape of the stator and rotor to optimize the performance of the motor,and in the optimization design process,it is often only a single-objective optimization.In the process of optimization,the modification of the structure parameters can only improve one target,and affect other important performance targets,so that the optimization efficiency is low,and a lot of calculation time is required.Through the research,it is found that the motor with dual-stator has the characteristics such as fast response,high overload capacity and high material utilization and so on.In this thesis,the dual-stator switched reluctance motor is selected to be optimized.In order to improve the optimal efficiency,reduce the computation time and cost,in this thesis,from the surrogate model,a multi-objective differential evolution algorithm with an adaptive opposition-based learning based on block point Kriging surrogate model is developed,the surrogate model is used to replace the simulation calculation,so that the running times of the finite element analysis simulation calculation are reduced,thus the optimization efficiency is improved.At the same time,the proposed algorithmis varified by the test functions and the TEAM 25 problem.The main work of this thesis are as follows:Firstly,the principle of the differential evolution algorithm is introduced,and a differential evolution algorithm with an adaptive opposition-based learning is proposed to improve the diversity of the population,in order to avoid premature convergence and jump out of local optimum,based on programming software,the optimization algorithm is programed,and it is verified through the test function.Secondly,the principle of block Kriging model is introduced,and on the basis of it,the method of optimal frontier parallel adding points is proposed to improve the block Kriging model,and its accuracy is verified by test functions and TEAM 25 optimization problem.Thirdly,the model of the dual-stator switched reluctance motor is established,and the sensitivity of the tooth width for outer stator pole width,rotor outer teeth pole width and rotor inner teeth pole width is analyzed,and the factors that have great influence on the torque smoothing coefficient and average torque of the dual-stator switched reluctance motor are determined,thus the optimal variables are determined.Finally,the program of the proposed-algorithm is written based on programming software.The test function and TEAM 25 problem are used to verify the performance of the improved optimization algorithm,and it is applied to the optimal design of a dual-stator switched reluctance motor,the improved algorithm is combined with practice.