Processo automático de restabelecimento de energia em sistemas de distribuição modelado através de sistemas multiagentes

Abstract

Based on research that demonstrates the fragility of current distribution systems, there is a growing interest of regulators, researchers and utility companies in studies seeking possible solutions to events that lead to shutdowns in these systems. In the occurrence of a fault in the distribution network, the protection system is triggered and the supply branch is separated into two parts: upstream and downstream. The portion located upstream are the energized loads while the downstream are the non-energized ones. In this situation, the network is in a state of fault, so it is important that the utilities restore the out-of-service loads, preferably in a short period of time. In this context, utilities and other interested agencies are continually seeking improvements in services, adopting Smart Grids concepts, such as encouraging research development and acquiring technologies with automated properties. In this sense, it is possible to notice the constant evolution and the strong investment driven by the Smart Grid concept, which brings technological solutions for modernization, giving rise to a certain expectation regarding future networks. In view of these observations, the present work aims to propose a restoration strategy based on Multiagent Systems to help in the search for a solution to the restoration problem in Distribution Systems. The model addresses the sequential stages of energy restauration in order to define, through the analysis of technical reports and operational guidelines, the rules, actions, and operating environment of each agent. This model considers the application of a strategic plan for energy restoration with the possibility of load shedding, which addresses the mapping of out-of-service areas, the assessment of the supply capacity of each feeder and the maximization of loads restored through of switching. The presented model is applied to a typical distribution network through simulations, and the obtained results are compared with similiar works, based on evaluation parameters including minimum voltages in the load buses, power losses, switching actions and the restored loads.