Continuous Model Predictive Control of Current-fed Half-Bridge Isolated DC/DC Converter

Abstract

This paper focuses on the control of the snubberless naturally-clamped soft-switching current-fed halfbridge isolated DC/DC converter. The average state-space model for the converter is established and linearized. Adopting the discrete-time linear predictive model, a continuous control set model predictive controller (CCS-MPC) with fixed switching frequency is firstly proposed for this converter control to improve the converter dynamic performance. Meanwhile, in order to solve the model-mismatch to achieve offset-free tracking, the extended-Kalman observer is designed to obtain the voltage error and the integrated error for dynamically compensating the current reference calculation as a feedforward compensator. Finally, the proposed control strategy is validated by the simulation of MATLAB/Simulink, and the proposed approach can stabilize the output voltage with good steady-state performance and strong robustness against the input voltage and load variations.