基于卡尔曼滤波的永磁同步电动机位置检测

Position Detection of Permanent Magnetic Synchronous Motor Based on Kalman Filter

研究了一种基于卡尔曼滤波的永磁同步电动机位置检测方案。在永磁同步电动机的矢量控制中,转子位置的检测精度对于控制性能十分关键。低成本的应用场合中,常采用3个空间互差120°的开关型霍尔来检测当前转子位置,但是霍尔安装误差、磁场分布不均匀等原因均会造成霍尔传感器之间的不一致,进而产生位置检测偏差。针对该问题,介绍了一种基于卡尔曼滤波的检测方案,结合霍尔的测量值,对真实的转速和位置进行观测。同时考虑了角加速度,对加减速段进行处理,使之数学模型更贴合实际。仿真和实验表明,所介绍的位置检测方案能够有效提高检测精度。

A method of position detection of permanent magnetic synchronous motor （PMSM） based on Kalman filter was proposed. The accuracy of rotor position detection is very important for the vector control performance of PMSM. In low cost applications, three binary-type Hall sensors arranged 120° apart were used to detect the rotor position. However, installation error and non-uniform flux distribution could lead to the inconformity of Hall sensors, and then position detection error was generated. To tackle this problem, a detection scheme based on Kalman filter was proposed. The actua/ speed and position were observed combined with measured value by Hall sensors. Meanwhile, the accretion period was specially deal with, so that the model was more aligned with the reality. In the end, experimental results show that the proposed scheme can improve the position detection accuracy.