Aplicação de um algoritmo genético para otimização do roteamento de veículos coletivos em rede

Abstract

Genetic algorithms are used in a wide range of optimization problems, especially in combinatorial problems, where the search space is, in many occasions, too large for exact methods to achieve optimal solutions. This class of problems bear great importance for mathematical models that reflect real world situations, such as vehicle routing. The goal of this work is to utilize a genetic algorithm to optimize networked linked capacitated vehicles, in this configuration, all passengers and drivers positions are known in every moment. To make this optimization possible, the problem was mathematically modeled using the Capacitated Vehicle Routing Problem (CVRP) as inspiration with alterations reflecting the nature of the network linked capacitated vehicles. After that, this work also compares the obtained performance to a non-optimized route, so that the gain from using this method can be observed.