Suppression of periodic speed disturbances in permanent magnet synchronous motors based on sliding adaptive filters

Abstract To suppress periodic speed disturbances in permanent magnet synchronous motors, this paper proposes a novel adaptive interference cancellation method that combines the Sliding Discrete Fourier Transform (SDFT) with the Least-Mean-Square (LMS) algorithm. It not only estimates the fundamental frequency of periodic disturbances but also automatically updates parameters to adapt to changes in disturbance frequency, thereby enhancing the control system’s resistance to disturbances. This approach utilizes an adaptive notch filter to replace the traditional low-pass filter, thereby reducing amplitude and phase losses in the control system signals. Additionally, by improving the fixed window of the LMS algorithm to a sliding window, the real-time performance of signal transmission in the control loop is enhanced. This paper designs a load torque observer for real-time monitoring of periodic load disturbances and inputs the disturbance-containing signal into the SDFT. The SDFT algorithm is utilized to estimate the disturbance frequency, which the adaptive notch filter then uses to suppress synchronous disturbances in both the motor control system and the output current. Simulations conducted in the MATLAB/Simulink environment show that the proposed method effectively suppresses periodic speed disturbances in the motor.