A low complexity multi-step progressive optimization virtual vector model predictive control strategy for grid connected converters

There is a problem of large output current ripple in the finite control set model predictive control of a grid-connected converter. Thus a low-complexity multi-step progressive optimization virtual vector model predictive control strategy is proposed. First, this method quantitatively evaluates the voltage vector error of each sector by constructing the voltage error equation for each sector. Secondly, virtual vectors are designed at the positions of maximum voltage errors in each sector to minimize current control errors. Then, the voltage vector errors of each sector are further quantified and analyzed by the proposed multi-step progressive optimization method, and the virtual vectors are designed at the positions of the largest voltage error, further reducing the control error. The multiple optimizations are used to determine optimal virtual vectors, reducing the current ripples efficiently. Finally, to alleviate the computational burden, a simplified sector determination search strategy is devised. Each major sector is subdivided into six smaller sectors, thereby reducing the number of candidate voltage vectors and enhancing the system’s dynamic response speed. The effectiveness of the proposed method is validated by experiment. © 2024 Power System Protection and Control Press. All rights reserved.