Two-Stage Optimization-Based Model Predictive Control of 5L-ANPC Converter-Fed PMSM Drives

Abstract

Balancing the dc-link voltage and three flying-capacitor (FC) voltages while maintaining very fast control of the motor is always a challenging task in five-level active neutral-point-clamped (5L-ANPC) converter-fed permanent-magnet synchronous motor drives, especially when the FC is small. Besides, the fundamental-frequency operation of the low-frequency cell (LFC) is also required to reduce the losses of the converter. This article proposed a two-stage optimization-based model predictive control (MPC) scheme to achieve these goals. In the first stage, the switching states of the LFC are selected based on the sign of the desired output voltage through a deadbeat approach for its fundamental-frequency operation; in the second stage, the optimal duty cycles for the high-frequency cell are calculated by the multiple-vector MPC and generated by phase-shifted pulsewidth modulation for its fixed-switching frequency operation. The FC and dc-link capacitor voltage balance can be managed by the flexibility of the inherent redundancy in the 5L-ANPC converter. Instead of 512, only six switching vectors must be considered, which significantly simplifies the computational burden of MPC. A thorough experimental evaluation has been conducted to validate the effectiveness of the proposed MPC method.