The Research On Parameter Optimization Strategy Of PMSM Sensorless Control

The high-performance operation of permanent magnet synchronous motor(PMSM)can’t be separated from the rotor position.The sensors installed to detect the rotor position,due to their volume,size,cost and service life,restrict the application of PMSM in more fields.In order to solve these problems,using different algorithms to build digital sensors instead of mechanical sensors to realize sensorless control has become one of the key contents in the field of motor research to broaden the scope of PMSM and save costs.The traditional PI regulator is widely used in the sensorless control.The parameters of the motor will change in bad operating conditions or special working conditions.However,the traditional PI regulator is easily affected by the changes of motor parameters and working conditions and can’t meet the precise control requirements of PMSM,so the parameters of PI regulator optimization method is also very important in the study of sensorless control.At the same time,many PI parameter optimization methods of PMSM are mostly for the parameter tuning of PI regulator of speed and current link,which in rotor position observer is relatively less.This paper analyzes the classification method of PMSM,the performance of various motors,the mathematical model of PMSM and the control strategy of PMSM.Adopting the method of high-frequency square wave signal injection,an optimization method of rotor position observer based on the combination of time multiplied absolute error(ITAE)criterion and error back propagation(BP)neural network is proposed.First of all,the mathematical model of PMSM is established,and its transformation in each coordinate system is also discussed.Finally,the high-frequency square wave signal injection method is used to track the rotor position and speed information.Secondly,this paper studies the error integration criterion.After analyzing and comparing the regulation characteristics of several error integration criteria,the ITAE Criterion is used to optimize the PI parameters of the rotor position observer.Finally,the neural network control theory is studied and the rotor position observer parameters of BP neural network with relatively simple structure,less parameters to be determined offline and strong generalization ability are used for further optimization.With the help of MATLAB simulation software to build PMSM simulation experiment model,the optimization effect of the above optimization method is compared with the traditional method.