An improved ant colony optimization and its application to vehicle routing problem with time windows

The ant colony optimization (ACO), inspired from the foraging behavior of ant species, is a swarm intelligence algorithm for solving hard combinatorial optimization problems. The algorithm, however, has the weaknesses of premature convergence and low search speed, which greatly hinder its application. To improve the performance of the algorithm, a hybrid ant colony optimization (HACO) is presented in this paper. By adjusting pheromone approach and introducing a disaster operator, the HACO prevents the search process from getting trapped in the local optimal solution. Then, by taking the candidate list into consideration and combining the ACO with the saving algorithm and λ-interchange mechanism, the HACO improves the convergence speed. Furthermore, this paper gives a guideline on how to adjust the parameters to achieve the good performance of the algorithm. Finally, the HACO is applied to solve the vehicle routing problem with time windows. The effectiveness of the HACO on solving combinatorial optimization problems is validated by comparing the computational results with those previously presented in the literature.