Research on Multi-Vector Cascaded Model Predictive Control of Induction Motors

Abstract

When using single vector model predictive control, the direction of output voltage in each control cycle is limited. Therefore, there is pulsation and a quite large harmonic distortion in stator current. In this paper, a novel cascade multi-vector model predictive control strategy is proposed. Firstly, the model predictive controller is used to replace the traditional linear controller in the external speed loop, and a three-vector model predictive controller is designed in the internal current loop. The amplitude and direction of output voltage vector are adjustable which improves the performance of the current control. As multi-vector model predictive control will increase the switching frequency of the system, we construct a cost function with switching frequency constraints to solve this issue. Finally, the proposed control strategy is verified on the asynchronous motor experiment set that it can improve the current tracking performance and effectively reduce the switching frequency of the inverter while maintaining the speed loop with good dynamic and static performance.