The improvement of ice transportation routing for minimizing the transportation cost by particle swarm optimization

Abstract

This paper attempts to improve the efficiency of ice transportation which is the one of difficult and challenging problem in the field of vehicle routing problems (VRP). Since the restriction of the ice quality if there is too much time spending in transportation would result in the melting of ice, the maximum duration of a route with soft delivery time window and the variant of vehicles with different capacity constraints must be considered. Therefore, the purpose of this paper is to provide the decision support tool for the ice transportation problem, based on optimization techniques. A well-known meta-heuristic called Particle Swarm
Optimization (PSO) is brought to develop an efficient algorithm. The objective function is the minimization of the total cost which consists of 1) a fixed cost of driver’s labor, 2) fuel cost of the truck and3) soft time window penalties. In order to avoid being trapped into a local optimum, a re-initialization strategy is used to improve the possibility of finding an optimal solution in the searching space. The results obtained from this study show that the proposed PSO is not only useful for reducing the transportation cost when compared to the firm’s current practice and other meta-heuristic techniques for all problem instances. Furthermore, the heuristic models developed in this study should prove beneficial to other, similar businesses in Thailand and
other interested parties.