Efficient Two-Vector-Based Sequential Model Predictive Control for IM Drives
Tongzhen Liu, Xuliang Yao, Jingfang Wang, Chenwei Ma
Abstract
In this article, an efficient two-vector-based sequential model predictive control (ETV-SMPC) method is proposed for induction motor (IM) drives. It aims to provide an effective scheme for SMPC with extended control set. A fast selection mechanism for the candidate set is presented to generate feasible combinations. The proposed method reduces the computational cost by excluding a considerable portion of the redundant candidates. Due to the flexible sequential evaluation structure, the cost functions for evaluating torque and flux errors are developed in different forms, effectively improving the evaluation performance over that of conventional designs. By analyzing the two-vector control implementation in different scenarios, the torque evaluation stage is designed for the global optimization of torque control. Considering the tracking errors at the switching instant and the end of the control period, the flux error evaluation process is further optimized. In addition, candidates for the flux evaluation stage are adjusted based on the preset torque tolerance. As a result, the proposed ETV-SMPC exhibits better steady-state performance. Additionally, it does not require weighting factors in the cost function and the corresponding tuning work. The experimental results confirm its effectiveness.