Application of Genetic Algorithms to Solve the Multidepot Vehicle Routing Problem

This paper deals with the optimization of vehicle routing problem in which multiple depots, multiple customers, and multiple products are considered. Since the total traveling time is not always restrictive as a time window constraint, the objective regarded in this paper comprises not only the cost due to the total traveling distance, but also the cost due to the total traveling time. We propose to use a stochastic search technique called fuzzy logic guided genetic algorithms (FLGA) to solve the problem. The role of fuzzy logic is to dynamically adjust the crossover rate and mutation rate after ten consecutive generations. In order to demonstrate the effectiveness of FLGA, a number of benchmark problems are used to examine its search performance. Also, several search methods, branch and bound, standard GA (i.e., without the guide of fuzzy logic), simulated annealing, and tabu search, are adopted to compare with FLGA in randomly generated data sets. Simulation results show that FLGA outperforms other search methods in all of three various scenarios. Note to Practitioners-The multidepot vehicle routing problem is one of the common optimization problems in the logistics area. In a real-world environment, drivers choose the shortest route to reach a destination since they assume that it should take the shortest time to travel the shortest route. However, if some events such as traffic congestions, accidents happen in the shortest route, the traveling time spent on this route can be greater than that on the longer route. Thus, this paper considers not only the cost due to the total traveling distance, but also the cost due to the total traveling time, as two objectives. We propose a stochastic search technique called FLGA to solve the problem. Based on the promising computational results obtained in this paper, the proposed model and technique will be effective for industries to be applied in solving real-world problems.