Priority-Based Model Predictive Control Method for Driving Dual Induction Motors Fed by Five-Leg Inverter

Abstract

This article proposes a priority-based model predictive control (PMPC) method for a five-leg voltage source inverter (FLVSI) that independently drives dual three-phase induction motors (IMs). In a FLVSI operated by a MPC method, selecting an optimal future voltage vector among all 32 candidate switching states causes a high computational burden. However, by using the PMPC, low computation and current ripples can be achieved at the same time. In the PMPC method, two IMs are prioritized, and the priority of both IMs alternates every control period. For this property, the future voltage vector for the high-priority motor (HM) is selected first; that for the low-priority motor (LM) is selected next. In other words, whereas future switching states of both IMs are optimized by one cost function in the conventional MPCs, those of the HM and LM are optimized by their cost functions in PMPC. Therefore, HM and LM have eight and four candidate voltage vectors for cost minimization, respectively, which results in low computation complexity. Furthermore, current ripples are small because the HM is optimized based on all possible candidate voltage vectors, and both IMs are selected as the HM equally. The PMPC method is demonstrated by simulation and experiment results.