A new ant colony optimization with global exploring capability and rapid convergence

Ant colony optimization (ACO) is a meta-heuristic algorithm, and is widely applied in combinatorial optimization. To enhance the ACO´s global exploiting capability and convergence, a new pheromone update strategy is presented, which results in a gradually transition of the ant colony´s diversity, and an improved ACO algorithm called ACO+ is proposed. For a solution to the traveling salesman problem (TSP), a statistical model of traversed ants of sub-routes is introduced to rank the sub-routes, and an adaptive pheromone trails update mechanism is implemented, which integrates with the iteration-best pheromone update strategy. The algorithm can effectively combine the global exploring capability and convergence rate. Experiments show that the ACO+ has a good performance and robustness.