Optimization and improvement of Genetic Algorithms solving Traveling Salesman Problem

Abstract

Traveling salesman problem (TSP) is a typical NP-complete problem, of which the search space increases with the number of cities. Genetic algorithm (GA) is an efficient optimization algorithm characterized with explicit parallelism and robustness, applicable to TSP. In this paper, we compare the performance of the existing GAs in searching the solution for TSP and find a superior combination of crossover and mutation method. Then, the improvements in the cycle crossover and greedy cross-cycle crossover are proposed. Finallyl experimental results show that the new cycle crossover and greedy crossover algorithms perform much better than the original ones.