一种永磁电机无速度扩展卡尔曼算法  

A Speed-Free Extended Kalman Method for Permanent Magnet Motor

作　　者：方晓洁 黄宇柏 黄伟琼 陈玲 高淇 桑仲庆 FANG Xiao-jie;HUANG Yu-bo;HUANG Wei-qiong;CHEN Ling;GAO Qi;SANG Zhong-qing(State Grid Zhangzhou Power Supply Company,Zhangzhou 363000,China;Institute of Electrical Engineering and Automation Chemistry,Xiamen University of Technology,Xiameng 361021,China)

机构地区：[1]国网漳州供电公司,福建漳州363000 [2]厦门理工学院电气工程与自动化学院,福建厦门361021

出　　处：《电气传动自动化》2019年第4期13-17,共5页Electric Drive Automation

摘　　要：在实际生产应用中永磁同步电机的控制精度非常关键,在去除噪声的过程中,我们通常采用卡尔曼滤波,一般的卡尔曼无法满足要求,我们提出一种新型扩展卡尔曼滤波器。在电机的参数的实时运算中,并将参数不断更新进算法使得精确度大幅度提高,同时在传统扩展卡尔曼(EKF)的线性化过程中,泰勒近似只是在上一个周期的最优处完成,因此无法做到完美更新数据,降低了精确度。在此提出一种低阶串行双扩展卡尔曼滤波算法,通过仿真和实验比较了EKF和LSDEKFs对电机转速估计的误差,结果表明新型扩展卡尔曼滤波器具有更优的估计精度。In order to improve the accuracy of speed sensorless vector control system for permanent magnet synchronous motor,a new low-order extended series Kalman filter is proposed.By extending the unstable motor parameters to the system as the vectors of the state to be identified,the online real-time calculation of motor parameters is realized,and the obtained parameters are projected into the algorithm to realize the accurate identification of motor speed,while transmitting them.In the linearization process of EKF,Taylor approximation is only completed at the optimum point of the last cycle,so the data lags behind and the accuracy is reduced.In this paper,a low-order serial dual extended Kalman filter algorithm is proposed.The speed estimation errors of the traditional extended Kalman filter and the new extended series Kalman filter are compared by simulation and experiment when the motor parameters change.The results show that the new Kalman filter has higher speed estimation accuracy.

关 键 词：永磁同步电机 无速度传感器 低阶串行扩展卡尔曼滤波算法 状态估计 

分 类 号：TM28[一般工业技术—材料科学与工程]