12 - Advances in predictive control of DC microgrids

Abstract

It is well known that by definition microgrids need control and management at different levels to allow the inclusion of renewable energy sources. The most spread renewable energy technologies like solar photovoltaic (PV) correspond to direct current (DC) technologies. This fact allowed that DC microgrids have appeared to be a very promising technology, which may easily include PV-based distributed generators, energy storage systems, and modern loads. The latter has favored DC microgrids to become one of the key research areas from last years in academia and electrical industry. As DC microgrids also need to be controlled to operate properly, islanded, or connected to the utility grid, its control has risen as a very challenging aspect as these small-scale electric systems have to fulfill grid codes to be feasible to be applied technically and economically to modern power electric systems. On the other side, predictive control has proved to be a very practical and feasible control scheme for power converters that interface renewable energy-based distributed generators. In this chapter, we survey and describe these predictive control-based techniques applied to power converter-interfaced DC microgrids, with their mathematical expressions and key factors to make the reader understand their development and applications to DC microgrids. In this way, we explain finite control set model predictive control, modulated model predictive control, decentralized model predictive control, and hybrid finite control set model predictive control/deadbeat predictive control as primary level controllers in DC microgrid systems. Also, we describe predictive control-based schemes used in the energy management system as secondary level controllers for a DC microgrid, that is, model predictive-based self-adaptive inertia control and a centralized model predictive control scheme.