| dc.description.abstract | A novel strategy is proposed in this work for power system restoration. Due to the nature of the objective functions and constraints, the restoration problem should be modelled as a non-linear multi-objective optimization problem. This makes it hard to find suitable solutions for the problem. In this proposal, a multi-objective genetic algorithm, Strenght Pareto Evolutionary Algorithm (SPEA2), was implemented with the goal of performing scans in order to generate efficient unique solutions. To prove the efficiency of the proposed strategy, one test system, with 16 buses, was considered. The algorithm creates as result a decoded individual with sequential solutions, always recovering a big amount of system load. The algorithm was then applied to two real large systems offered by Cemig Distribution, one with 703 buses and another with 484 buses. The algorithm generates as a result individuals decoded over sequential solutions. A novelty presented in this paper is that the algorithm not only minimizes the load disconnected at the end of the set of maneuvers, but also minimizes disconnected after each iteration load, thus the decision-maker will make sure that the sequence of maneuvers is presented that retrieve the largest amount of charge. | |
