改进扩展卡尔曼滤波器的PMSM参数辨识

Improved Extended Kalman Filter Method for Parameter Identification of Permanent Magnet Synchronous Motor

针对传统扩展卡尔曼(EKF)辨识永磁同步电机(PMSM)参数时难以确定合适的系统与测量噪声矩阵和精度较差问题,文章提出一种粒子群算法(PSO)优化EKF的PMSM参数辨识方法。首先剖析EKF原理,建立EKF辨识模型,然后利用PSO自适应优化EKF的系统噪声矩阵和测量噪声矩阵,并根据EKF辨识模型设计出合适的适应度函数,从而使EKF获取更优的参数辨识效果。仿真结果表明,该方法能较好辨识PMSM的电感与磁链参数,比传统方法具有更好的辨识精度和速度。

To address the problem of difficulty in determining the appropriate system and measurement noise matrices and inferior precision in the traditional Extended Kalman Filter(EKF)identification of Permanent Magnet Synchronous Motor(PMSM)parameters,this paper proposes one PMSM parameter identification method of Particle Swarm Optimization algorithm(PSO)optimized EKF.The EKF principle is analyzed atfirst,an EKF identification model is established,and then the system noise matrix and measurement noise matrix of the EKF are optimized by PSO adaptively.Moreover,a suitable adaptation function is designed according to the EKF identification model,resulting in a better parameter identification effect of the EKF.The simulation results show that the method can better identify the inductance and magnetic chain parameters of the PMSM,and has better identification accuracy and speed than the traditional method.