Enhanced Ant Colony Optimization with Local Search

Abstract

The ant colony optimization (ACO) algorithm frequently gets trapped around local optimum solutions and does not approach the global optimum solution of vehicle routing problem. This work attempts to remedy this drawback by modifying the ant system (AS) algorithm, an instance of ACO. The modification includes a new method of route construction, new weight for improving the pheromone density of each route, and introduction of SA to improve the quality of the solutions from local search. It is called an enhanced ant colony optimization with local search or EACOL. A performance test was performed on EACOL on 10 standard datasets comparing it to those of the ant system algorithm and elitist ant system, and the results show that the proposed algorithm performs better than these two algorithms on these datasets.