An Adaptive Finite-Control-Set Model Predictive Control to Reduce Computational Complexity for Multilevel Converters

Multilevel converters have been used in various applications. In practice, the common-mode voltage (CMV) requires suppression to improve system reliability. In this case, the finite-control-set model predictive control methods (FCS-MPC) are appropriate for achieving current tracking and CMV suppression to operate converters normally. However, conventional FCS-MPC (CFCS-MPC) faces the challenges of reducing its complexity when applied to multilevel converters. In order to address the above issue, an adaptive finite-control-set model predictive control (AFCS-MPC) method is proposed in this article. AFCS-MPC reveals the relationship between current tracking and CMV suppression through geometric analysis, thus eliminating the tedious loop calculation in FCS-MPC, thereby eliminating the need for traversing the lookup table and avoiding tedious iterative calculations. AFCS-MPC can be applied to different multilevel topologies with a simple adjustment of two parameters. No prestored data, like state combinations or vector coordinates, is needed, whereas they are all calculated online. AFCS-MPC does not involve complex calculations such as trigonometric functions or root signs. The theoretical analysis clearly indicates that the computational burden of the AFCS-MPC is basically constant for any multilevel converter. The effectiveness of the AFCS-MPC method is finally verified by simulation and experiment.