Performance enhancement using durational model predictive control combined with backstepping control and disturbance observer for electrical drives

Model predictive control (MPC) is widely used in a cascaded structure for electrical drives, in which, the MPC is applied as inner-loop for current regulating while another controller such as proportional-integral (PI) is applied as outer-loop for speed regulating. However, the benefits of MPC for multiple targets optimizing simultaneously are not sufficiently utilized. Moreover, characteristics of these drives rely on the outer-loop controller due to the cascaded strategy, and they are parameter-sensitive. A durational MPC is proposed in this paper to optimize the speed and currents of permanent-magnet synchronous machine (PMSM) simultaneously, and it is combined with backstepping control (BSC) and disturbance observer to improve the disturbance rejection ability and parameter robustness. The nonlinear characteristics of drives are considered, and the voltage reference is constructed by combining the optimal voltage vector and virtual voltage target produced by MPC and BSC, respectively. The proposed method takes advantages of both MPC and BSC and thus provides significant performance enhancement compared with the conventional PI-base cascaded MPC. Simulation and experiment for the PMSM drives with the proposed method are conducted to prove its stability, feasibility, and efficiency.