Robustness Improvement of Deadbeat Model Predictive Control for Five-phase PMSM Drives

Abstract

Deadbeat (DB) model predictive control (MPC) is one of the most advanced control technologies, which has the advantages of fast dynamic response, simple implementation, and low computational-burden. However, DB-MPC still suffers from the problem of parameters mismatch sensitivity, which may deteriorate the performance under the parameters mismatch or uncertainties. This paper proposes a recursive least squares (RLS)-based estimation strategy to improve the robustness of the DB MPC for five-phase permanent magnet synchronous machine (PMSM) drives. With the aid of RLS estimation, the whole PMSM’s parameters can be precisely estimated, hence it can be applied on interior PMSM, as well as surface PMSM. The proposed control algorithm is compared with the conventional DB MPC, and the latest robust MPC algorithm, which is based on the online current difference update mechanism. The simulation and hardware-in-the-loop (HIL) experimental results verify the effectiveness of proposed algorithm.