基于扩展卡尔曼滤波器的电机参数辨识算法

Electromagnetic Parameter Identification Algorithm of AC Motor Based on Extended Kalman Filter

准确的电机参数是无速度传感器交流调速系统设计的前提。运行过程中,由于温度、频率和磁通的影响,电机参数通常会发生变化,特别是,转子电阻和磁链电感的变化直接关系到磁链和转速的辨识精度,影响驱动系统性能。本文通过测量定子电压和电流,基于扩展卡尔曼滤波器(EKF),研究了一种新颖算法,通过将两个基于扩展卡尔曼滤波器模型有机结合,协同工作,实现了对转子电阻、励磁电感、转子磁链和转速在线辨识。仿真和实验结果表明这种新颖的电机参数辨识算法具有较小的辨识误差,可以满足无速度传感器交流调速系统对电机参数准确性的要求。

It is the precondition for AC drive system to determine the electromagnetic parameters of induction machine exactly. But in the running process of the motor, the electrical parameters will change, that maybe result from the effect of temperature, frequency, and flux. Especially, variations of stator resistance and magnetizing inductance are not only relevant to identification error of flux and speed directly, but also impact of the driven system performance. By measuring stator voltage and current, a novel algorithm is presented based on the serial operation of two extend Kalman filter （EKF） in this paper.This approach will achieve rotor stator resistance, magnetizing inductance, rotor flux, speed as well as high-precision identifications online. Experimental and simulation results show that the developed algorithm leads to smaller system error than the traditional one and can meet the precision requirement in speed sensor-less vector control system of induction motor.