Three-Level Inverter-PMSM Model Predictive Current Control Based on the Extended Control Set

Abstract

In the neutral point clamped (NPC) three-level inverter-permanent magnet synchronous motor system, traditional model predictive current control (MPCC) uses the system predictive model to traverse the 27 basic voltage vectors, to achieve the d-q axis current component and neutral point voltage of the multi-objective optimal control. Finite control set model predictive control predicts the state change of the control target at future moments based on a finite number of switching states of the inverter. The control principle of this method is simple and easy to implement, but the control effectiveness of this control strategy is limited because only one basic vector can be selected as the optimum output per control period. In this paper, a model predictive current control strategy based on an extended control set (ECS-MPCC) is proposed, which can improve the control performance of the system by extending the control set to select multiple vectors in a single control period compared to the traditional strategy. In addition, to address the disadvantage of extending virtual space vectors leading to an increase in computation, this paper proposes a fast search method for optimal vector based on region reduction. The proposed method avoids the optimization process traversing all virtual space vectors, thus enabling a fast search for the optimal vector. The experimental results show that the proposed control strategy has good steady-state and dynamic performance.