Agents-Based Service Restoration in Electrical Secondary Distribution Network

Service restoration (SR) is one of the fundamental functionalities of distribution management system during fault management process. Several approaches have been devised to solve the SR problem on distribution network including centralized and distributed approaches. Most of the existing studies focused on the service restoration in the medium voltage network with very few focusing on the secondary distribution network (SDN). In most countries including Tanzania, the service restoration is accomplished manually through relying on operational experiences, rated capacity of the transformers and peak hour demand for decision making. This study aimed at designing distributed algorithm based on multi-agent system for SR in the SDN. The study has been conducted through intensive literature review together with focus group discussion with key stakeholders from utility company, study visits to the Tanzania SDN in Dar es Salaam region and interviews with the technical people. SDN with three transformers rated at 315, 200 and 100 kVA have been chosen as the pilot site for designing the distributed algorithm. The designed algorithm includes designing of multi-agent system, objective functions to be optimized and design for the equipment specifications and power system network topology to support SR. Four agents, namely control agent, grid agent, load agent and switch agent, have been found to be optimal for the SR process. The designed restoration process mainly focuses on load transfer to the nearby transformer and load shedding. The future work will focus on real implementation of the designed algorithm and consideration of the integration of the renewable distributed generations.