Multi-agent technology for self-healing shipboard power systems

Shipboard electric power systems (SPS) are dynamic, complex systems that require continuous monitoring and fast controls to operate in a reliable and secure state. A self-healing system for SPS was developed using common programming languages to arm SPS to monitor and respond to threats and natural system failures. This self-healing system includes features for activating control solutions before weapon hits or impending equipment failures, and after damage-induced or natural faults occur to continuously maintain system operation in an optimal state. This paper discusses a new implementation of the self-healing system using multi-agent system (MAS) technology. A potential implementation of one function in the self-healing system using MAS is presented. It is concluded that the characteristics of MAS provide a natural framework for developing a self-healing methodology for SPS that is automated and enhances fight-through survivability