An Improved Heuristic Ant Colony Algorithm for Capacitated Vehicle Routing Problems

In view of the slow convergence and the low quality of ant colony algorithm for capacitated vehicle routing problems(CVRP), an improved ant colony algorithm is proposed. Inspired by the ant colony system and the rank-based ant system, the algorithm employs the pheromone update strategy, not only strengthening the usage of the best solutions of each iteration, but also avoiding falling into a local optimum. With the use of the basic method by a certain percentage and construction of a path with pushforward insertion heuristic(PFIH)method, the search space is expanded. Using a hybrid local search operator, it enhances the capacity of local search. The experimental results show that the algorithm can greatly reduce the vehicle operating costs and has fast convergence speed, which proves that the algorithm is effective.