| Summary: | 碩士 === 國立高雄第一科技大學 === 運輸倉儲營運所 === 93 === This paper analyzes a different case of the Vehicle Routing Problem (VRP) that includes the time-dependent link weights. In addition, a modified algorithm was proposed and it had been applied onto a real road network. This case is more realistic to the real world. For the logistics and supply chain management, how to dispatch vehicles to serve customers and decide the order of services are needed to reconsider under this situation for more efficient delivery and better customer service quality. Moreover, most traditional VRP assume straight-line connection between any two nodes rather than taking path selection into account. It is also discussed in this paper. The Ant Colony Optimization (ACO) algorithm is used in this paper for solving this VRP. It simulates the ant society, which is able to approach a good feasible solution. In addition, Multiple Ant Colony System (MACS) is employed for solving two goals simultaneously, which are to minimize the number of vehicle and the total travel time. A new pheromone-updating scheme was proposed and proved to be feasible for solving VRPTW with continuous speed-time relationship function link weight. Furthermore, an application prototype is developed to illustrate the solution of this algorithm. This application demonstrates a user interface (UI) of solving VRPTW and using the map of Kaohsiung city as study object.
|
|---|
