Chattering phenomenon analysis and suppression of sliding mode observer in PMSM sensorless control

Spectrum analysis of PMSM back EMF sliding mode observer (SMO) based on windows FFT shows spectrum range of chattering signal is large and the amplitude is high. The result can't be used directly and must be filtered. But the signal changes after passing the low band filter and the compensation is difficult. This paper presents an extended SMO, of which the state variables are current and flux. The observed flux is composed of the real flux and the filtered chattering signal. The spectrum analysis shows the influence of chattering signal can be ignored, and the motor angle can be calculated from the observed flux. The low band filter is omitted, and the phase delay is avoided. The influence of speed estimation error to flux observation is analyzed, and feedback matrix is calculated based on the analysis. Simulation and experimental results show that the proposed method can precisely estimate the position and speed of the rotor, and the system has good dynamic and static performances.