An improved ant colony algorithm is designed. The given algorithm combines ant colony system (ACS) and
max-min ant system (MMAS), and introduces the factors of time windows width and waiting time for service into state transfer rules. In order to avoid the algorithm trapping into local optimum, different pheromone evaporating rules are used under different stages. 2-opt which is an intra-optimizing method in a road and 2-opt* which is an inter-optimizing method between roads are used to improve the optimizing solution received from each iteration. Experiment results show that the algorithm is better than genetic algorithm and Tabu search algorithm on aspects of solving effects and operation effectiveness.