Ant supervised by PSO and 2-Opt algorithm, AS-PSO-2Opt, applied to Traveling Salesman Problem

AS-PSO-2Opt is a new enhancement of the AS-PSO method. In the classical AS-PSO, the Ant heuristic is used to optimize the tour length of a Traveling Salesman Problem, TSP, and PSO is applied to optimize three parameters of ACO, (α, β, ρ). The AS-PSO-2Opt consider a post processing resuming path redundancy, helping to improve local solutions and to decrease the probability of falling in local minimum. Applied to TSP, the method allowed retrieving a valuable path solution and a set of fitted parameters for ACO. The performance of the AS-PSO-2Opt is tested on nine different TSP test benches. Experimental results based on a statistical analysis showed that the new proposal performs better than key state of art methods using Genetic algorithm, Neural Network and ACO algorithm. The AS-PSO-2Opt performs better than close related methods such as PSO-ACO-3Opt [9] and ACO with ABC [19] for various test benches.