Distributed Energy Optimization in MAS-based Microgrids using Asynchronous ADMM

This paper proposes a scalable distributed energy management system for multi-agent system (MAS)-based microgrids under communication delay. In the proposed approach, the centralized convex optimization problem of a microgrid with distributed energy resources is decomposed into sub-problems solved iteratively by the respective agents in a distributive manner using privacy-preserving asynchronous alternating direction method of multipliers (ADMM) algorithm. In this algorithm, each agent accumulates the sub-system updates from local controllers and only shares partial information with the adjacent agents while minimizing the aggregated operational cost of the microgrid in a cooperative mode. The numerical result shows that the algorithm can reach the convergence very fast under various conditions. The optimality of this algorithm was analyzed briefly and its effectiveness was compared with the synchronous ADMM algorithm.