A Study of Solving Traveling Salesman Problem with Genetic Algorithm

Abstract

Travelling salesman problem is one of the most important problems in the optimization area. To solve the problem of traveling salesman, the GA (genetic algorithm) can be seen as an appropriate method. In the genetic algorithm, there are many parameters needing to be set in advance. To figure out the effects of parameters in GA, we conducted experiments with different parameters and compared the performance. In addition, we try to improve the algorithm in aspects of crossing and mutating. The improved GA obtained a better performance, showing the proposed improvement is efficient.