Initial position estimation for surface permanent magnet synchronous motors based on magnetic saturation effect

In this paper, the magnetic saturation model for surface permanent magnet synchronous motors (SPMSMs) is adopted to detect the initial rotor position. The model is also verified by the transient finite element analysis (FEA) method. The proposed algorithm is based on the injection of high-frequency (HF) rotating voltage vector, which produces the HF current response. Then this weak HF current is filtered and amplified significantly via a high pass and resonance filter which can also remove the DC components effectively. After that, an improved demodulation method is introduced to extract the useful information including the negative-sequence DC component and the iron core saturation evaluation function. The rotor direction is then estimated primarily via a phase-locked loop (PLL) with the extracted negative-sequence component and the magnetic polarity is furthermore confirmed via injecting a DC voltage vector at the estimated rotor direction and comparing the saturation evaluation function before and after. Overall the rotor position is estimated with the rotor direction and the magnetic polarity. The proposed initial position algorithm is validated by the 3.6 kW hardware drive system as well as on the Matlab/Simulink simulation platform using the suggested magnetic saturation model. It is proved to be simple, reliable, and with high precision.