Model Predictive Control Of Three-Phase Voltage Source PWM Rectifier Under Unbalanced Grid Conditions

PWM rectifiers are widely used in grid-connected power conversion systems,however,unbalanced grids may lead to a reduction in power quality and affect the normal operation of the converter.In recent years,Model Predictive Power Control（MPPC）has received a high level of interest due to its conceptual simplicity and no required for complex control parameter design.To improve the control performance of three-phase voltage-source PWM rectifier（VSR）in unbalanced grids,this paper uses the MPPC strategies to develop the study.Firstly,the operating principle of the three-phase VSR is introduced in terms of its topology and operating modes.Then its mathematical model is developed in the balanced and unbalanced grids respectively.Then,the implementation principle of finite control set MPPC（FCS-MPPC）is presented.The FCS-MPPC discretizes the rectifier control model and then selects the optimal switching state at the next moment to drive the switching tubes based on the principle of minimisation of the cost function.To improve the control performance of PWM rectifiers under unbalanced grids,two MPPC control strategies are proposed in this paper.Scheme I makes the MPPC suitable for grid voltage imbalance by adding compensation power to the reference power;Scheme II uses extended instantaneous power theory to achieve both constant active power and grid-side sinusoidal currents control under unbalanced grids,which no need for complex sequences decomposition and power compensation algorithms.The conventional FCS-MPPC has a high harmonic content of grid side currents due to the unstable switching frequency,for which an optimized MPPC（OSS-MPPC）is proposed.The multi-vector synthesis method is used and the selected vector switching sequence is reorganized for optimisation and the action time of each vector is calculated,which enhanced the system control performance.In addition,a finite control set model predictive optimisation control strategy for LCL filter-based VSR is proposed.The multivariable control feature of MPC is exploited by constructing a single cost function to simultaneously control the grid side current vector,the filter capacitor voltage vector and the rectifier side current vector（MPC-i₁i₂u_c）,which without the need for active damping of the state variable feedback,and then using the constructed cost function J to select the optimal switching state at the next moment.Finally,a three-phase VSR test prototype based on single-L filtering was built,and the experimental verification of the proposed MPPC strategies were completed.The results show that under unbalanced grids,the two equivalent optimisation schemes proposed in this paper both have a significant improvement on the system control performance.However,option II has a simpler control structure as no power compensation is required.By restructuring and optimising the switching sequences,OSS-MPPC solves the problem of variable switching frequency of FCS-MPPC and further reduces the grid side currents THD value.