Model Predictive Nearest Level Control (MP-NLC) Method for 9-Level Converter With LC Filter

Multilevel Converter (MLC) design and control have been interesting topics in the last few decades. Due to the increased number of semiconductors in such topologies, the control problem becomes complex. Model Predictive Control (MPC) has been introduced in the 1980s and owing to significant advancements in processing power nowadays it is considered a viable option to control power converters. Fast dynamic response, simplicity, and the ability to include system nonlinearities and constraints in the cost function make MPC an interesting alternative for conventional linear controllers. In this paper, conventional Finite Control Set Model Predictive Control (FCS-MPC) has been improved by limiting the viable switching options based on the reference value. The performance of the proposed Model Predictive Nearest Level Control (MP-NLC) method is compared to the conventional FCS-MPC. The proposed method is applied to a 9-level inverter. To investigate the performance of the proposed MP-NLC method, simulations were carried out under linear and nonlinear loads. The presented results verify the fast dynamic response, and robustness of the proposed method. In addition, parameter mismatch analyses have been carried out on the proposed method as well. Based on the presented results, by using the proposed method total efficiency and specific power of the system are increased.