A new Ant colony optimization metaheuristic based on pheromone guided local search instead of constructive approach

Abstract

This paper proposes a new algorithm based on a new idea inspired by the Ant Colony Optimization (ACO) metaheuristic. The goal of the proposed algorithm is to increase the performance of the traditional ACO algorithms by changing the classic role played by ants. In fact, ants are not anymore used to construct solutions but to improve the quality of a population of solutions using a specific local search strategy. Thus the new algorithm, called Ant-PLS, is a population based local search algorithm which uses the stored knowledge in pheromone trails to guide the ant when choosing its neighboring solution from a dynamic candidate list. Ant-PLS is applied on the symmetric traveling salesman problem. The experimental results show the effectiveness of the Ant-PLS when applied on large benchmark instances of TSP. In fact Ant-PLS results are significantly better than the compared ACO algorithms.