Model Predictive Direct Speed Control With Novel Cost Function for SMPMSM Drives

Abstract

The model predictive direct speed control (MPDSC) can improve the speed dynamic performance effectively. However, due to the large difference between the mechanical and electrical time constants, short prediction could result in large current ripples at steady state, which becomes a critical problem. Besides, the conventional MPDSC methods have not deeply investigated the online-adjustable weighting factor design and the problem of <i>i_d</i> transient oscillation. In this article, a model predictive direct speed control with novel cost function is proposed. In the proposed method, two different <i>q</i>-axis current references are generated, one to improve the speed dynamics, and the other to ensure the low current ripples. The two current tracking are assigned with different weighting factors unified in one cost function. In order to determine which current tracking to play a major role under different operations, the speed error is designed as weighting factor. With the proposed cost function and designed weighting factor, the high speed dynamics and low steady current ripples can be achieved simultaneously. Finally, the effectiveness of proposed method is verified by experiments.