基于强跟踪容积卡尔曼滤波的永磁同步电机转子位置估计

Estimation of Rotor Position of PMSM Based on Strong Tracking Cubature Kalman Filter

对于电动汽车内置式永磁同步电机(IPMSM)驱动系统,转子位置的精度在高性能无传感器矢量控制中起到极其重要的作用。当电机在运行过程中受到外界干扰和系统状态突变时,传统的容积卡尔曼滤波(CKF)算法的动态响应较差,导致对转子位置的跟踪能力下降,估计精度降低,甚至引起滤波器发散。因此,本文采用强跟踪容积卡尔曼滤波(STCKF)算法,在传统的容积卡尔曼滤波的基础上引入了强跟踪滤波器(STF),进而动态改善容积卡尔曼滤波算法的估计精度和跟踪能力。在Matlab/Simulink中对改进的转子位置估计算法进行仿真分析,并且进行测功机台架实验。实验结果表明:强跟踪容积卡尔曼滤波算法响应快,跟踪能力强,估计精度相比于传统的容积卡尔曼滤波算法提高19%。

For the interior permanent magnet synchronous motor(IPMSM)drive system of electric vehicles,the accuracy of the rotor position plays an extremely important role in high performance sensorless vector control.When the permanent magnet synchronous motor is disturbed and the system state is abrupted during operation,the dynamic response of the conventional cubature Kalman filter(CKF)algorithm is poor.Then tracking ability and estimation accuracy for the rotor position are reduced,and the filter is even diverged.Therefore,this paper adopted strong tracking cubature Kalman filter(STCKF)algorithm.The improved algorithm introduced a strong tracking filter(STF)based on the traditional cubature Kalman filter,then dynamically improved the estimation accuracy and tracking ability of the cubature Kalman filter algorithm.The improved rotor position estimation algorithm was simulated and analyzed in Matlab/Simulink,and the dynamometer bench experiment was carried out.The experimental results show that the strong tracking cubature Kalman filter algorithm has fast response and strong tracking ability in medium and high speed,and the estimation accuracy is 19%higher than the traditional cubature Kalman filter algorithm.