Um operador de cruzamento baseado na preservação dos blocos de construção

Abstract

n this work, a genetic crossover operator based on the preservation of the building blocks that exist between the parent chromosomes has been introduced. This operator is used in place of the genetic crossover operator of a cutoff point in Holland's genetic algorithm, where the preservation of the building blocks that exist in the parents in their children occurs, with the purpose of solving optimization problems. The genetic algorithm with the proposed operator, is compared with the same algorithm considering one cross cut point and two cut points. These three operators were tested on two optimization problems, traveling salesman and OneMax. To evaluate the performance of the proposed crossover operator compared to the genetic one-cutting point cross operator, where two types of analysis will be performed: convergence of the genetic algorithm and quality of the chromosomes found. At the end of the tests, the results obtained by the proposed operator with one cutoff point, and with two points, and the genetic crossover operator of one cutoff point of Holland's genetic algorithm were analyzed.