Author(s)

Jianping Luo, Xia Li, Min-Rong Chen

College of Information Engineering, Shenzhen University, Shenzhen, 518060, China.

In this work, we present a multi-phase hybrid algorithm based on clustering to solve the multi-depots vehicle routing problem (MDVRP). The proposed algorithm initially adopts K-means algorithm to execute the clustering analyses, which take the depots as the centroids of the clusters, for the all customers of MDVRP, then implements the local depth search using the Shuffled Frog Leaping Algorithm (SFLA) for every cluster, and then globally re-adjusts the solutions, i.e., rectifies positions of all frogs by the extremal optimization (EO). The processes will continue until the convergence criterions are satisfied. The results of experiments have shown that the proposed algorithm possesses outstanding performance to solve the MDVRP.

Combinatorial Optimization; Vehicle Routing Problem; Shuffled Frog Leaping Algorithm

J. Luo, X. Li and M. Chen, "Multi-Phase Meta-Heuristic for Multi-Depots Vehicle Routing Problem," Journal of Software Engineering and Applications, Vol. 6 No. 3B, 2013, pp. 82-86. doi: 10.4236/jsea.2013.63B018.