A best-path-updating information-guided ant colony optimization algorithm

Abstract The ant colony optimization (ACO) algorithm is a type of classical swarm intelligence algorithm that is especially suitable for combinatorial optimization problems. To further improve the convergence speed without affecting the solution quality, in this paper, a novel strengthened pheromone update mechanism is designed that strengthens the pheromone on the edges, which had never been done before, utilizing dynamic information to perform path optimization. In addition, to enhance the global search capability, a novel pheromone-smoothing mechanism is designed to reinitialize the pheromone matrix when the ACO algorithm's search process approaches a defined stagnation state. The improved algorithm is analyzed and tested on a set of benchmark test cases. The experimental results show that the improved ant colony optimization algorithm performs better than compared algorithms in terms of both the diversity of the solutions obtained and convergence speed.