Switching loss optimization based on model predictive control for grid-connected inverter

A method based on the model predictive control is proposed to decrease the switching loss of single-phase grid-connected inverter. It detects the current direction based on the discrete state model of the inverter to select the switching state preliminarily and to block the states of some devices, takes the tracking current difference as the index function and substitutes the selected switching state into the index function to search the optimal switching state resulting in the minimum index function. A loss reduction function is then introduced to evaluate the current control error, if the error does not exceed the limit, the switching state output remains, otherwise changes the switching state output according to the present minimum of index function to significantly reduce the system switching frequency. The proposed control method is simple in principle and easy to implement the digital control. Compared with traditional control methods, the system is more robust and no PI parameters are needed. Simulative results show that, the controller tracks the grid-connection current well and the switching frequency is greatly reduced. ©, 2015, Electric Power Automation Equipment Press. All right reserved.