Multi-Objective Optimization Three-Vector-Based Model Predictive Current Control with Adaline Online Parameter Identification for PMSM Drives

Although the conventional three-vector model predictive current control (C-M3PCC) has improved the accuracy compared with the conventional single-vector MPCC, the calculation amount increases exponentially. In this paper, an improved three-vector MPCC control strategy based on Adaline parameter identification is proposed. Firstly, the motor's dynamic and steady states are identified and segmented to simultaneously enhance their control performance, and the sector selection method is improved to reduce calculation time. In addition, Adaline neural network parameter identification aims to enhance the system's robustness when system parameters mismatch. Finally, the zero vector selection method is improved to reduce the switch loss. This paper verifies that the proposed method can achieve superior precision and robustness control effect at lower switching frequencies.