Implementación de la metaheurística FEPSO GIST mediante procesamiento paralelo: Aplicación al problema de balance de fases en sistemas de distribución en baja tensión

Abstract

This paper describes the design and implementation of the FEPSO GIST (Fuzzy Particle Swarm Optimization with Global-Individual Star Topology) Metaheuristic, under a paradigm of parallel computing platform using the MPI (Message Passing Interface). The model is applied to the solution of Multiobjective Unbalanced Load Problem in Low Voltage Electric Distribution Systems (ULP LV EDS). First, the ULP LV EDS is presented, which has no solution by classical optimization methods. Then, the framework over the heuristics methods has a wide range of application and the parallel computation is the main approach to accelerate the results of algorithms that require a more powerful computing process respect of sequential paradigm, is described. Finally, the FEPSO GIST Metaheuristic design, in this framework, is presented, and its implementation MIP is applied on a real ULP LV EDS.