Optimal Decentralized Coordination of Sources in Islanded and Grid-connected Microgrids
Abstract
Rising climate change concerns in recent years have instigated the emergence of sustainable sources to reduce dependence on high-emission, isolated bulk generation systems. The microgrid framework relies on integrating these distributed energy resources (DERs) to achieve regional energy independence that leads to a reliable and environment-friendly power grid. In this work, highly granular and decentralized coordination schemes are proposed that will enable fast computation of source dispatch set-points, thereby appropriately accounting for frequent changes in regional load-supply configuration of a microgrid. The mathematical models utilized in the study sufficiently represent the steady-state electrical interdependencies and feasibility limits in islanded or grid-connected operation modes. Applying alternating direction method of multipliers (ADMM) algorithm, the coordination process is transformed into a decentralized multi-agent problem involving minimal information exchange between subsystems. To overcome non-convexities typically present in optimization problem of microgrids, the two distinct convex relaxation techniques utilized are: 1) Linearization and 2) S-procedure. In separate coordination schemes, the former lends the advantage of extremely fast computation speed, while the latter exploits the hidden convexity in the decomposed coordination problem to deliver solutions with superior feasibility guarantee. Finally, the convergence, feasibility and scalability of the proposed coordination techniques are assessed with simulation studies performed on realistic microgrid parameters and several IEEE test systems.