Flatness-Based Decentralized Control of Bidirectional Interlink Power Converters in Grid-Connected Hybrid Microgrids Using Adaptive High-Gain PI-Observer

Abstract

This article studies the control of bidirectional interlink power converters (BILPCs) in hybrid microgrids (HMGs). Here, the BILPCs interconnect a dc microgrid to an ac microgrid in a grid-connected HMG. Thanks to the flatness property of differential equations of BILPCs, the control method is able to provide desirable output characteristics and robustness against unmodeled dynamics and unknown inputs. First, the flat models of single BILPC and parallel-connected BILPCs are determined. Second, to approximate the unknown inputs and disturbances such as changes in the controlled output voltages of BILPCs, an adaptive high-gain proportional-integral observer is designed for BILPCs. Based on a linear quadratic regulation approach, the high gains of this observer change dynamically during the approximation process, to achieve the best result. Third, the approximated parameters are implemented in the design process of sliding mode control-based decentralized controllers for BILPCs. The simulations confirm the capability of the new method considering the dynamic response and power transferring control capability among the ac and dc microgrids.