Optimization-Based Initial Position Detection for Interior Permanent Magnet Synchronous Motor (IPMSM) With High-Frequency Injection

Abstract

Accurate initial position is crucial for the start-up or operation performance of interior permanent magnet synchronous motor (IPMSM). This article proposes an optimization-based initial position detection scheme. Different from conventional scheme, a new high-frequency (HF) voltage equation is developed and it is used to establish a cost function as the position is an unknown variable. Then, according to the convex cost function, position estimation can be achieved through optimization method. The proposed method does not need a low-pass filter (LPF) or bandpass filter (BPF). It could achieve a faster convergence speed and adapt to distorted higher injected signals. Since signal modulation and filters are no longer required in the proposed scheme, the cumbersome process to adjust the filter parameters becomes unnecessary. Another, for the ambiguity of $\pi $ , sliding discrete Fourier transform (SDFT) and phase locked loop (PLL) are combined to extract the amplitude of current harmonic quickly and accurately. The proposed scheme has been verified experimentally on a laboratory test.