Performance analysis of a new algorithm for power distribution system reconfiguration

Distribution system reconfiguration (DSR) is a multi-objective, non-linear problem. This paper demonstrates the performance and effectiveness of a new, fast, non-dominated sorting genetic algorithm (FNSGA) for the purpose of solving the DSR problem by satisfying all objectives simultaneously with a relatively small number of generations and relatively short computation time. The objectives of the problem are to minimize real power losses and improve the voltage profile and load balancing index with minimum switching operations. Instead of generating several ranks from the non-dominated set of solutions, this algorithm deals with only one rank; then the most suitable solution is chosen according to the operator´s wishes. If there is no preference and all objectives have the same degree of importance, the best solution is determined by simply considering the sum of the normalized objective values. Different crossover operators and rates, population sizes, load variation, storing of previously considered topologies, and different initial topologies are applied to a widely studied test system. The results show the efficiency of this algorithm in a wide range of parameters such as crossover rate, crossover operator, population sizes, generation number and loading conditions.