Online Inductance Identification of PMSM Based on High Frequency Signal Injection into Virtual Axis

Dianguo Xu,Shaobo Liu,Kaiji Zhang,Gaolin Wang,Guancheng Pan,Qiwei Wang
DOI: https://doi.org/10.1109/ICEMS56177.2022.9982799
2022-11-29
Abstract:The voltage equation-based method is the most commonly applied inductance identification method. However, the effectiveness of this method is affected by the parameter coupling and the rank deficient. To solve this problem, an online inductance identification method is proposed based on the virtual axis high-frequency (HF) sinusoidal voltage signal injection strategy. According to the virtual axis HF equivalent impedance model which includes the position information between the virtual axis and the dq axes, the dq-axis inductances can be calculated by extracting the HF signal of the virtual axis. The proposed method is able to make sure that the identified parameters decouples from other motor parameters and operating conditions and the method is immune to sensorless control position error. The dq-axis inductance can be identified without rank deficient at full operation conditions, especially at low current and speed condition. The proposed method is validated on a 2. 2kW PMSM drive.
Physics,Engineering
What problem does this paper attempt to address?