Grid Voltage Sensorless Model-Based Predictive Power Control of PWM Rectifiers Based on Sliding Mode Virtual Flux Observer

In this paper, a grid voltage sensorless model predictive control is proposed based on a sliding mode virtual flux observer (SMVFO). The proposed SMVFO shows good inherent filtering capacity, and thus there is no high-frequency chattering problem. In addition, the proposed SMVFO is designed based on the closed-loop current estimation. Not only is DC-drift issue solved but also dynamic response is enhanced when compared with the prior open-loop virtual flux observer. To verify the effectiveness of the presented SMVFO, it is further integrated into finite control set-model predictive control (FCS-MPC) for pulse width modulator (PWM) rectifiers. The whole control algorithm features simplicity and improved cost-effectiveness due to the absence of modulator and grid voltage sensors. As the SMVFO can predict current at the next sampling instant while estimating virtual flux accurately, the proposed SMVFO assisted FCS-MPC is comparable to its counterpart using measured grid voltage. The simulation and experimental tests were carried out on a two-level voltage source PWM rectifier to validate the effectiveness of the proposed method.