Robust Model Predictive Current Control Based on Inductance Sliding Mode Observer

Abstract:
For model predictive control (MPC) strategy, the outstanding feature of MPC method is that control performance depends on the accuracy of model parameters. In order to improve the parameter robustness of MPC, this paper introduces an optimization strategy. Firstly, the influence of different model parameter mismatches in MPC method on control performance is analyzed. And it is proved that inductance and flux linkage have great influence on predictive control performance. Therefore, this paper introduces the corresponding improvement measure, in which, a flux linkage processing method is adopted to replace the flux linkage parameters in the prediction model with the resistance and inductance information. This means that the flux linkage is removed from the prediction model. Then, for the optimization of inductance parameters, the sliding mode observer is adopted to observe the inductance disturbance value, and the compensated inductance value is substituted into the current prediction model. Finally, the experiment results prove the effectiveness of the proposed method.