Improved multi-objective evolutionary algorithm in subpopulation tables with features from NSGA-II for the service restoration problem
Leandro T. Marques, José Paulo Ramos Fernandes, João B. A. London Jr · 2019
Problems in electric power supply cause economic losses and affects people's lives. To reduce these impacts, Distribution Systems (DSs) operators must have service restoration plans, which must respect a series of constraints while reaching different objectives to properly restore the system. An improved version of the Multi-Objective Evolutionary Algorithm (MOEA) based on both the Non-Dominated Sorting Genetic Algorithm - II (NSGAII) and in the MOEA in Subpopulation Tables is proposed to get better Pareto fronts. In order to generate faster implemented and cheaper plans, the ability to prioritize switching operations in Remotely Controlled Switches (RCSs) is kept. Proposed MOEA was compared with four MOEAs from literature in a real and very large-scale DS in two different scenarios and outperformed them all in the tests conditions.