A Discrete-Time Algorithm for Real Time Energy Management in DC Microgrids

Abstract

A microgrid will often be characterized by very low embedded system inertia when operating fully islanded. In this state, management of the embedded stored energy in the microgrid over timescales of milliseconds is of critical importance as even short periods of mismatch between generation and consumption will cause voltage collapse on the network. This article proposes a discrete-time (sampling-based) energy management algorithm that uses the rapid controllability of power-electronics interfaced loads to manage the internal stored energy of loads with the aim of minimizing the use of bulk storage and dispatchable generation. Fast interconverter communication is used to balance storage demand across the network and to dispatch generation. A lab-scale microgrid with 10 nodes and a sampling time of 10 ms is used to investigate the effectiveness of the proposed energy management algorithm.