Model Predictive Control for Isolated DC/DC Power Converters with Transformer Peak Current Shaving

摘要

In this paper a Model Predictive Control, suitable for isolated DC/DC power converters and featuring optimization on converter operations, is proposed. The control is designed based on a specific a current-fed DC/DC converter topology, named Active-Bridge-Active-Clamp converter. This topology is particularly suitable for batteries interface, utilizing its interleaved structure in order to achieve an effective current ripple cancellation at the low voltage terminals. However, the increased complexity of the ABAC converter requires a specific modulation implementation and a detailed converter modeling. For such reasons, the operating principle of the ABAC converter is firstly introduced and mathematical models of the ABAC are developed in this paper. Subsequently, a Model Predictive Control is proposed and implemented aiming to achieve terminal current regulation, improve dynamic performance and reduce current stress in a wide DC voltage operating range, whilst maintaining a fixed switching frequency and a reduced prediction horizon. Simulation results for a 10kW ABAC converter are provided to validate the theoretical claims.