MAS-based Distributed Load Restoration in Resilient Networked DC Microgrid Systems

Extreme events such as faults, cyber-physical attacks, and natural disasters have a widespread impact on the electrical system infrastructure, leading to extensive power outages or blackouts. These events can cause temporary or permanent damages to the power systems and seriously compromise the system's resiliency. A quick restoration strategy for the systems' critical loads after extreme events is crucial to ensure resilient operation. This paper presents a fast load restoration (LR) approach for multi-agent-system-based networked DC microgrid systems with distributed energy resources. The overall restoration problem is designed with a weighted sum optimization technique that consists of operational costs, load pickup costs, and optimum power-flow-related costs. It was solved in a distributed manner by applying Alternating Direction Method of Multipliers (ADMM) algorithm through multi-agent communications. The effectiveness of the developed LR technique is verified by applying it to a networked MG system of 4 single-bus DC MGs.