Genetic algorithms applied to the design of large power distribution systems

Abstract

This paper presents the application of a new genetic algorithm for the optimal design of large distribution systems, solving the optimal sizing and locating problems of feeders and substations using the corresponding fixed costs as well as the true nonlinear variable costs. It can be also applied to single stage or multistage distribution designs. The genetic algorithm has been tested with real size distribution systems achieving optimal designs in reasonable CPU times compared with respect to the dimensions of such distribution systems. On the other hand, these distribution systems present significantly larger sizes than the ones frequently found in the technical literature about the optimal distribution planning. Furthermore, original operators of the genetic algorithm have been developed in order to obtain global optimal solutions, or very close ones to them. An integer codification of the genetic algorithm has also been used to include several relevant design aspects in the distribution network optimization.