Research On Multi-vector Model Predictive Control Of Permanent Magnet Synchronous Motor Based On Visualization Analysis

In recent years,with the increasing demand of society for green environmental protection,energy conservation and emission reduction,new energy power generation technology and new energy electric vehicle technology have both developed vigorously.Permanent magnet synchronous motor(PMSM),as a new type of drive motor with high efficiency and energy saving,has received more and more attention.Based on the theory of model predictive control,a multi-vector model predictive current control method for PMSM is studied in this paper,which solves the effects of different parameters mismatch on the model predictive control system of PMSM under different cost functions.Meanwhile,it also theoretically solves the effectiveness of the multi-vector model predictive control strategy for PMSM based on three-phase six-switch and three-phase four-switch inverters.Firstly,a visualization analysis method is proposed for the model parameters mismatch on model predictive control of PMSM,and the effects of different parameters mismatch on model predictive control system of PMSM with different cost functions are studied.The visualization method is used to analyze the effects of different parameters mismatch on the optimal voltage vector selection under different cost functions,laying a foundation for the following analysis of multi-vector model predictive control under different cost functions.Experimental studies have verified the feasibility and effectiveness of the proposed method.Secondly,due to the problem of large current ripples in conventional single-vector model predictive control of PMSM,a multi-vector model predictive control method of PMSM is adopted to reduce current ripples.However,the multi-vector model predictive control method lacks strict theoretical basis.Therefore,in this paper,a visualization analysis method is proposed and applies it to analyze multi-vector model predictive control under different cost functions,further proving the effectiveness of multi-vector model predictive control.Meanwhile,based on visualization analysis methods,it is found that conventional double-vector and triple-vector model predictive control strategies cannot achieve optimal control.Based on this,a new hybrid multi-vector model predictive control strategy is proposed.This strategy improves current control performance by selecting the optimal voltage vector from 18 alternative voltage vector combinations.Comparative experimental studies verify the effectiveness of the proposed model predictive control strategy.Finally,the multi-vector model predictive control is applied to three-phase fourswitch inverter driven PMSM systems to solve the problem of large current ripples.Meanwhile,in this paper,a visualization analysis method considering DC voltage pulsation is presented to evaluate the effectiveness of the conventional multi-vector model predictive control strategies for PMSM system driven by three-phase four-switch inverter.However,the analysis results show that the conventional triple-vector model predictive control strategy is not completely superior to the conventional double-vector model predictive control strategy.Therefore,a hybrid multi-vector model predictive control strategy is also proposed in three-phase four-switch inverter systems,which further reduces current control error and ripple by integrating conventional double-vector and triple-vector model predictive control strategies.Comparative experimental studies verify the effectiveness of the proposed model predictive control strategy.