Research On Robust Design And Multi-objective Optimization For Permanent Magnet Motor Considering Parameter Perturbation

As the core component of electric vehicles,the performance of drive motor directly decides the performance of the electric vehicles.And optimization,as the effective way to improve the motor performance,has been widely concerned in the motor design field.However,facing the strict performance requirement of the high quality drive motor,the performance objectives tend to be diversity,and with the constantly emerging and improve of the novel motor topology,the motor geometry design parameters also gradually increase.In addition,in the whole life design and development cycle of the motor,there are many uncertainty factors exist in the motor manufacturing and assembling process,which will further promote the optimization design of the motor towards to the direction of high dimentional and reliable development.Hence,how to integrate the uncertain factors in the process of motor manufacturing and assembling into the optimization design of motor,and effectively realize the comprehensive trade-off improvement of motor performance,is the research hotspot and hot direction in the field of motor,and has important research significance.Based on this,this paper take an interior permanent magnet motor as research target,proposed a robust design and multi-objective optimization method considering parameter perturbation for permanent magnet motor.Based on the initial design of the motor and the in-depth analysis of the multi design parameter model,the idea of variable dimensionality reduction is introduced to realize the effective screening of key design parameters.Then,the multi-objective optimization and robust multi-objective optimization considering parameter perturbation are studied respectively to provide beneficial support for the better comprehensive performance of the motor.The specific research work include the following aspects:1.The multi-variable model of the permanent magnet motor and parameter reduction design method are studied,the influence degree of design parameters on the motor performance is analyzed,and the importance of parameter reduction design for motor optimization is revealed.Taking an interior permanent magnet motor as an optimization design example,the dimensionality of design parameters is reduced,and the main design parameters that dominate the motor performance are determined.2.Taking torque,torque ripple and core loss as optimization objectives,a kriging interpolation model used to replace actual finite element simulation is constructed.Based on this model,a deterministic multi-objective optimization design method based on non-dominated sorting genetic algorithm is performed on the motor,and the geometric size of the motor under deterministic multi-objective optimization method is determined.3.Introduced the concept of motor robust optimization design considering parameter disturbance,and proposed a calculation method for the success rate of one-time machining of the motor considering parameter disturbance,and based on this,proposed a multi-objective robust optimization design method considering parameter disturbance for permanent magnet motor,and through the establishment of the Kriging proxy model and Monte Carlo analysis,based on the non-dominated sorting algorithm,the optimal design scheme of the motor considering the parameter disturbance is determined,including the deterministic motor geometric size and the allowable fluctuations of design parameters.4.The different optimization results obtained by the two optimization design methods are compared and analyzed.The torque,torque ripple,core loss of the motor,and the performance fluctuation considering parameter disturbance and the qualified rate of the motor are compared and analyzed from two aspects of electromagnetic performance and the robust performance of the motor under parameter disturbance.It is verified that the optimization design method proposed in this paper is superior to the deterministic multi-objective optimization design method based on non-dominated genetic algorithm under the condition that the design requirements of permanent magnet motor are satisfied.5.The prototype was processed and an experimental test platform was built.Through the measurement and analysis of the back EMF,steady-state torque and harmonic distribution of the experimental prototype,the rationality and correctness of the motor design and optimization process were verified.