부산대학교 도서관

In cascaded H-bridge (CHB) converters, each of the H-bridge carriers of the traditional modulated model predictive control (MPC) is synchronized. Therefore, the steady-state and transient performances are weak in high power applications that exhibit a low switching frequency. The traditional observer-based MPC uses estimation strategies to replace the prediction model under parametric uncertainties. However, there is no effective approach for the uncertainties in the open-loop optimal control law, which leads to a marked decrease in control performance in the presence of high uncertainties. For high power CHB converters, a robust closed-loop MPC using a linear matrix inequality is designed in this study to achieve robust current tracking. The augmented state is asymptotically decreased by the state feedback control law in the terminal elliptical invariant sets. Thus, the worst performance of the model mismatch is minimized. This study also proposes a phase-shifted modulated MPC to improve both the steady-state and transient performances. The voltage of the CHB converter is considered as a whole to implement the proposed method. The optimal duty cycle of each H-bridge, based on the piecewise strategy and voltage-second balanced rule, is implemented by a phase-shifted modulator. Better control performance with shorter response delays, faster tracking speed, and lower overshoot are achieved with a similar switching frequency. The effectiveness of the proposed method is verified by experimental results.