Three-Vector-Based Low-Complexity Model Predictive Direct Power Control Strategy for PWM Rectifier Without Voltage Sensors

Abstract

A voltage sensorless control of low-complexity model predictive direct power control (LC-MPDPC) for pulsewidth modulation (PWM) rectifier is proposed. The conventional LC-MPDPC adopts one or two voltage vectors during one control period, which achieve good steady-state performance and quick dynamic response. In addition, based on the mathematical model of the real system, the conventional method only requires one prediction to find the optimal voltage vector, which reduces the control complexity and computational burden. However, the use of one or two vectors during one sampling interval presents abundant current harmonics and high power ripples, and the switching frequency is variable. In order to solve these problems while preserving all the advantages of the conventional LC-MPDPC, this paper presents a novel control scheme, with the aim of operating at a constant switching frequency and obtaining an excellent steady-state performance at a low switching frequency. The proposed method is based on an optimal application of three voltage vectors in a symmetrical way, which takes advantage of advanced PWM techniques. Furthermore, the virtual flux-based control scheme is introduced to achieve voltage sensorless control. The proposed strategy is compared with the conventional MPDPC methods and its effectiveness is confirmed by both simulation and experimental results from a three-phase PWM rectifier under 1000-W operation condition.