A novel ODV crossover operator-based genetic algorithms for traveling salesman problem

Abstract

The genetic algorithm is a popular meta-heuristic optimization technique whose performance depends on the quality of the initial population and the crossover operator used to manipulate the individuals to obtain the final optimal solution. It is evident that when similar principle is followed for population seeding and crossover operators, it can enhance the speed of convergence and the quality of final individuals. The recent and popular population seeding technique for combinatorial genetic algorithm is ordered distance vector-based population seeding which works best with respect to convergence rate and diversity. However, the technique could not achieve the zero error rate convergence for the large-sized test instances. Thus, in this paper, an ordered distance vector-based crossover operator is proposed that exclusively exploits the advantages of individuals’ generated using the same initialization methods to attain the complete convergence, particularly for most of the large-sized test instances considered. One of the famous combinatorial problems of traveling salesman problem obtained from standard library is chosen as the testbed. From the experimental results, the proposed genetic algorithm model outshines the other existing and popular working genetic algorithm models in the literature.