Novel Explicit Model Predictive Control Strategy For Boost Converters Based on State-space Averaging Method

A novel explicit model predictive control strategy is proposed for DC-DC converters in this study. Firstly, the state-space models of boost converter are established, both on-state and off-state respectively. By characteristic analysis of state-space functions, the control target is reconfigured as a linear parametric-varying (LPV) model with time-variant state matrices. Towards such target, then an explicit model predictive controller (MPC) is proposed in order to enhance transition dynamics. A novel prediction model is designed by utilizing of Tylor series. Moreover, estimated average states are given as one of the objective variables in cost function by measurement of state-space averaging (SSA) method. Consequently, the computational load of boost converter control system is alleviated adequately. At the end, two numerical simulations of voltage tracking are performed, one in waveform of slope and the other is sinusoidal. The results show remarkable performances of rapid response without any steady-state errors.