Hybrid Model Analysis and Constructive Algorithms in the Performance of a Genetic Algorithm for the Planning of an Electric Power Transmission System

It has become more evident that in complex problems the performance of a genetic algorithm depends on the initial population, specially on its quality and diversity. In this paper four strategies are presented which, used concurrently allow to find quality and diversified topologies to set up an initial population to be used in a genetic algorithm to solve the problem of planning and expansion of electrical energy transmission systems. Also, experimental tests on real systems are presented to show the proposal's performance. This strategy to generate quality and diversified initial solutions can also be used in other algorithms such as simulated annealing, tabu search, GRASP, etc.