Evolutionary algorithm based on discrete ITO process for travelling salesman problems

Abstract

Since dozens years ago, various metaheuristic methods, such as genetic algorithm, ant colony algorithms, have been successfully applied to combinational optimization problem. However, as one of the members, ITO algorithm has only been employed in continuous optimization, it needs further design for combinational optimization problem. In this paper, a discrete ITO algorithm inspired by ITO stochastic process is proposed for travelling salesman problems (TSPs). Some key operators, such as move operator, wave operator, are redesigned to adapt to combinational optimization. Moreover, the performance of ITO algorithm in different parameter selections and the maintenance of population diversity information are also studied. By combining local search methods (such as 2-opt and LK-opt) with ITO algorithm, our computational results of the TSP problems show that ITO algorithm is currently one of the best-performing algorithms for these problems.