Algoritmo genético: um estudo comparativo dos métodos de seleção aplicado a um problema de otimização de carregamento em caminhão

Abstract

Genetic algorithms simulate the natural selection proposed by Darwin, being used in several areas of knowledge. They are famous for solving combinatorial problems, such as the traveling salesman problem and the knapsack problem. In this work, a variant of the knapsack problem is treated with a genetic algorithm, which seeks the best way to transport a combination of products that does not exceed the limit of a truck and maximizes the sales value (the sum of the price of all products in this load) without any item being repeated. In addition to finding the best solution to the proposed problem, the secondary objective is also to discuss the difference in the performance of the algorithm when using the roulette wheel selection method and the tournament selection method. The algorithm was coded in Python, as in addition to being a widely used language, it is also highly capable of working with lists. At the end of this study, it was concluded that the algorithm coded with the tournament selection method is more efficient than the roulette wheel method, confirming the literature in this area of research.