A Many-objective Ant Colony Optimization applied to the Traveling Salesman Problem

Abstract

Evolutionary algorithms present performance drawbacks when applied to Many-objective Optimization Problems (MaOPs). In this work, a novel approach based on Ant Colony Optimization theory (ACO), denominated ACO λ base-p algorithm, is proposed in order to handle Manyobjective instances of the well-known Traveling Salesman Problem (TSP). The proposed algorithm was applied to several Many-objective TSP instances, verifying the quality of the experimental results using the Hypervolume metric. A comparison with other state-of-the-art Multi Objective ACO algorithms as MAS, M3AS and MOACS as well as NSGA2 evolutionary algorithm was made, verifying that the best experimental results were obtained when the proposed algorithm was used, proving a good applicability to MaOPs.