Position Error Signal Extraction Strategy Based on High-Frequency Signal Injection Method

In traditional high-frequency (HF) signal injection methods, band-pass filters (BPF) are generally constructed for filtering carrier frequency signals that commonly are much larger than the injection frequency and fundamental frequency signals that are much smaller than the injection frequency. After signal modulation, the error signals with the position of the mover are obtained through low-pass filters (LPF) to realize real-time detection estimation of electric motor position. However, using filters not only causes time delay, but also reduces the available bandwidth of the position detection structure, which can significantly impact the anti-interference and stability of the detection results. For this reason, this paper optimizes the extraction strategies for the mover position signal under the traditional HF pulsating injection and the HF square wave injection, reduces the number of filters, expands the bandwidth of the position detection structure, and improves the anti-interference and stability of the sensorless control system. Finally, the superiority of the adopted strategies was verified through the RT-Lab semi-physical simulation platform.