An Ant Colony Optimization Memorizing Better Solutions (ACO-MBS) for Traveling Salesman Problem

Abstract

Ant Colony Optimization (ACO) is a population-based meta-heuristic method that mimics the foraging behavior of the ant colony in real life. The pheromone approach as the highlight method of the algorithm is the most effective factor in determining the moving of ants. Therefore, the problem of tuning the pheromone trail is an important topic for ACO that deserves attention. In this paper, a novel method which memorizes the solution costs and updates the pheromone trail according to the memorized costs is introduced for updating the pheromone trail in ACO. The performance of the proposed method was simulated on the Travelling Salesman Problem (TSP) and compared with the versions of ACO algorithm.