Distribution System Self-Healing Implementation using Decentralized IED-based Multi-Agent System

In this paper, local automatic switching plans, considering distribution network topology and hourly loading, are used to implement a self-healing strategy, which minimizes the switching operations to restore the interrupted customers within the shortest time interval. This self-healing implementation is done using decentralized multi-Agent system (MAS) where all local switching agents have awareness about the current network state and work cooperatively to reach the global purpose of the distribution network automatic restoration. We are proposing the utilization of intelligent electronic devices (IEDs) as the hosts of the local switching agents. The embedded code for the agent's autonomy, as well as the information exchange among agents are described, implemented and tested.