Sliding Mode Observer with Adaptive Parameter Estimation for Sensorless Control of IPMSM

To improve the observation accuracy and robustness of the sensorless control of an interior permanent magnet synchronous motor (IPMSM), a sliding mode observer based on the super twisting algorithm (STA-SMO) with adaptive parameters estimation control is proposed, as parameter mismatches are considered. First, the conventional sliding mode observer (CSMO) is analyzed. The conventional exponential approach law produces a large chattering phenomenon in the back EMF estimation, which causes a large observation error when filtering the chattering through the low-pass filter. Second, a high-order approach law of the super twisting algorithm is introduced to observe the rotor position and speed estimation, which uses the integral function to eliminate the chattering of the sliding mode. Third, an adaptive parameter estimation control (APEC) is presented to enhance the observation accuracy caused by parameter mismatches; the motor parameter adaptive law of the APEC is designed by Lyapunov's stability law. Finally, the proposed method not only reduces both the chattering and the low-pass filter, but it also enhances accuracy and robustness against parameter mismatches, as discussed through simulations and experiments.