Model predictive control of interleaved DC-DC boost converter with current compensation

Abstract

A cascade control scheme is constructed combing a model predictive control (MPC) scheme as a current-mode controller with an Extended Kalman observer to control an interleaved DC-DC boost converter under load and input voltage step change conditions. A state-space average model is established and linearized near the stable point for predicting the future dynamic behavior. The process of the solving optimization problems with constraints of cost function is converted to the quadratic programming problem. An Extended Kalman observer is designed to obtain the load resistance variation and the integrated voltage error which is used to dynamically adjust the current reference. The algorithm has been tested in simulation. The results show that compared with PI controller a good transient and steady-state responses in the presence of load variations and input voltage disturbances.