A New MPC-based Approach for Torque Ripple Reduction in BLDC Motor Drive

Abstract

In this paper, a new model predictive control (MPC) is proposed for brushless DC motor (BLDCM) to reduce the commutation torque ripple (CTR). The torque ripples generate vibration noise and reduce the efficiency. With purpose of minimizing the CTR of the BLDCM and considering the CTR sources, the proposed MPC scheme is designed by predicting the phase current and electromagnetic torque. The error square of predicted values of non-commutating current and electromagnetic torque, which are minimized in the cost function, determines the optimal switching states. The proposed MPC control is applied at commutation moments which is detected by analysis of Hall sensor signal. This control scheme is implemented on the traditional topology of the BLDCM driving system which facilitates the implementation. Considering a 210V-2000W BLDCM, the comparative analysis using the MATLAB/Simulink environment is carried out in terms of the CTR reduction, tracking of the reference current under low-speed, high-speed and load torque tracking. The key parameters’ responses of BLDCM illustrate the improvement of the CTR, fast-transient response and small steady-state errors by using the proposed MPC against the conventional PI-PWM.