基于衰减记忆卡尔曼滤波的无刷直流电机转子位置估计

摘要

Aiming at the problem that the Kalman filter is easy to diverge in estimating the rotor position of the brushless DC motor,a motor rotor position estimation method based on the memory attenuated Kalman filter (MAEKF) was designed to realize commutation control of the motor.The state equation of phase current,motor speed and rotor position angle as state variables was established.The goal of reducing the sensitivity of the motor control system to model deviation and cumulative error was achieved by optimizing the Kalman filter gain.The motor control system based on EKF was designed,and the rotor position angle and the sequence of the motor windings were given.Simulation and experimental results show that the motor can provide accurate commutation signal at no-load,load and load torque moment,the stable operation of the system was not appear divergence,the control system has strong robustness and feasibility.%针对卡尔曼滤波器在无刷直流电机转子位置估计中易发散的问题,设计了基于衰减记忆卡尔曼滤波器(MAEKF)的电机转子位置估计方法,实现电机的换相控制.建立了以相电流、电机转速和转子位置角为状态变量的状态方程,通过优化卡尔曼滤波器增益实现降低电机控制系统对模型偏差、累积误差等敏感程度的目标,设计了基于MAEKF的电机控制系统,给出了转子位置角和电机绕组导通次序.通过仿真和实验验证了电机在空载、带负载、负载转矩突变时刻均能提供准确的换相信号,系统稳定运行没有出现发散的问题,控制系统具有较强的鲁棒性和可行性.