Model Predictive Control for Voltage Regulation in Bidirectional Boost Converter

Abstract

In order to solve the problem of nonminimum phase behavior in DC-DC boost converters, multistep prediction techniques have been used in conventional model predictive control methods. However, this method suffers from a high computation burden since exponential numbers of switching states have to be evaluated. Moreover, it is difficult to design the input filter of the converter by using the conventional model predictive control method because of the variable switching frequency. To remove these shortcomings, this paper proposes an indirect model predictive control method in which the outer voltage loop is governed by a simple PI controller while the inner current loop is regulated by a model predictive controller. In comparison with the conventional model predictive control method, the proposed method realizes many outstanding features, such as being simple and computationally efficient with low computational burden. Besides, low cost operation is achieved since output current measurement is dispensable. Also, the proposed method provides a suitable duty ratio within a single step prediction horizon, which guarantees constant switching frequency. The effectiveness of the proposed control scheme is verified by Psim simulation.