A multiagent system based optimal load restoration strategy in distribution systems

Natural disasters pose a significant impact on the distribution infrastructures and often cause severe power outage including critical services. This paper proposes a new restoration approach to restore critical loads using distributed energy resources (DERs). Specifically, the critical load restoration problem, considering both the availability of DERs and the optimal restoration path, is formulated as a variation of the shortest path problem. The objective aims at maximizing the reliability of the restoration plan to decrease the possibility of postrestoration failures. A distributed approach framework embedded min/max-consensus, and bias min-consensus algorithms is developed, which only requires information exchange among neighbors based on a peer-to-peer communication protocol. The proposed distributed restoration approach is adaptive to topology changes and robust to unreliable communication network. Numerical results with various scenarios demonstrate the effectiveness of the proposed distributed restoration approach.