Software agents situated in primary distribution networks: A cooperative system for fault and power restoration management

In this paper, extended research upon the potentials of implementing distributed artificial intelligence technology to achieve high degrees of independency in distribution network protection and restoration processes is presented. The work that has already been done in the area of agent-based and/or knowledge-based applications and expert systems is briefly reviewed. The authors justify the need to distribute activities in contradiction to the centralized methodologies. A proper model of the real environment is introduced in order to define the designing parameters of a prototype agent entity, which is the part of a cooperative network-management system. The system's goal is to autonomously perform effective fault management upon medium-voltage power distribution lines. The structure of the agent entity is then described by means of the agent behaviors being implemented. The cooperative operations of the proposed system and its computer simulation are presented. Simulation results are being evaluated. Finally, general conclusive remarks are made.