Ultra-Local Model-Free Predictive Current Control for PMLSM drive systems with Moving Horizon Estimator

Abstract

Model-based predictive current control (MBPCC) methods have been applied in the field of motor control due to its better dynamic response and intuitive design, but the high performance of MBPCC are easy to be influenced by parameter mismatch which could lead to the deterioration of control performance. In this work, a model-free predictive current control method based on ultra-local model with moving horizon estimator (MFPCC-MHE) is proposed to overcome the influence of parameter mismatch on the MBPCC to improve the robustness of system and simplify the design of current loop. The total disturbance of system which contained known terms and unknown disturbance terms are obtained by MHE accurately. Simulation results show that the proposed MFPCC-MHE realize the effectively eliminate the influence of PMLSM parameter mismatch and has good robustness.