A low‐complexity FS‐MPDPC with extended voltage set for grid‐connected converters

Abstract

Abstract The conventional finite control set model predictive control (FS‐MPC) for converter control is a well‐studied area, but performance degradation due to the finite candidate vector set is still limiting its practical applications. Extending the voltage vector set using discrete space vector modulation has been proposed as a solution to overcome the limitations, but the brute‐force search inherent to FS‐MPC increases the computational complexity for a larger voltage set. This paper proposes a technique to alleviate the above issue by avoiding the brute‐force search that is being executed in FS‐MPC. The technique utilises the basics of direct‐power‐control theory to cut down the number of candidate voltage vectors applied in each cycle in the optimization problem. In this work, a design example having a voltage vector set of 37 elements is considered, and the proposed technique narrows down the search to eight optimal vectors. The proposed controller is specifically designed for active–reactive power control of a grid‐connected converter that interlinks an energy storage system to the grid. The system is modelled in MATLAB Simulink environment and simulations are carried out to analyse the performance in all four active–reactive bidirectional power flow modes. Results validate the performance of the controller, both in steady‐state and transient conditions. Further, the reduction in computational complexity due to the proposed algorithm is evaluated. It is observed that the number of computations was reduced approximately by 75% after applying the proposed algorithm for a system with a 37 voltage vector set.

Keywords

• computational complexity
• power grids
• predictive control
• reactive power control
• optimal control
• voltage control