Finite-Set Model Predictive Control for 17-Level Inverter with Reduced Number of Iterations in Photovoltaic Applications

Finite-set model predictive control (FS-MPC) algorithms provide excellent dynamic performance for the power electronics converters with the ability to easily include any non-linearities and constraints. However, the high number of iterations required for prediction and cost function evaluation, especially for multi-level converters, hampers the implementation of the FS-MPC techniques. In this paper, firstly, the deadbeat concept is utilized to directly obtain the reference voltage vector. Accordingly, the iterations required for prediction are eliminated. Secondly, the number of iterations for cost function evaluation is also reduced by selecting a certain number of voltage vectors based on the value of the reference voltage vector. Finally, experimental results using hardware in the loop (HIL) technology are given to validate the proposed FS-MPC technique.