Solving the traveling salesmen problem through genetic algorithm with new variation order crossover

Abstract

Genetic Algorithm (GAs) is used to solve optimization problems. It is depended on the selection operator, crossover and mutation rates. In this paper Roulette Wheel Selection (RWS) operator with different crossover & mutation probabilities, is used to solve well known optimization problem Traveling Salesmen Problem (TSP). We have compared the results of RWS with another selection method Stochastic Universal Selection(SUS), which demonstrate that the SUS is better for small number of cities; but as the number of cities increases RWS is far much better than SUS. We have also compared the results with a variation between mutation & crossover probability which concludes that mutation is more effective for decimal chromosome. We have proposed a new crossover operator which is variation of Order Crossover (OX) and found results are better than existing crossover operator.