부산대학교 도서관

In this paper, the concept of Finite Control Set Model Predictive Control (FCS-MPC) for a 17-level, single-phase switch ladder multilevel inverter (SP-SLMLI) connected to the grid is proposed. The current control of the inverter is a primal objective with the least computational effort. The inverter has 17-level output and hence the 17 control inputs. The inverter’s discrete model is employed to determine the future state of the control variable. In each sampling interval, there are 17 states to be checked, and the state that best minimizes the certain objective function defined in terms of the control variable is selected to be applied in the next sampling instant incorporating the controller delay due to calculations. Multilevel inverters with low switching loss and hence less power loss and suitable for medium and high power applications come with a substantial computational burden on the controller side. In order to stave off the calculation burden, conformal mapping has been introduced in the controller structure, and search space is reduced, thus requiring less time to search the optimal state per sampling interval. The idea can be extended to other Multilevel inverters with a higher number of switching levels.