| Summary: | 碩士 === 國立成功大學 === 交通管理學系碩博士班 === 94 === As the development of logistics, transportation or delivering cost, which accounts for substantial proportion of overall cost, could be reduced to enhance the competitiveness of the enterprise. Thus, the issue of minimizing transportation cost has drawn a lot of attention. Due to the advancement of Intelligent Transportation Systems (ITS), dispatching centers can arrange vehicle routes according to real-time information. The application of ITS technologies can improve deliver/pick-up services; however, the critical question is how to solve the Vehicle Routing Problems (VRP) under real-time information.
The Vehicle Routing Problems, belong to the class of NP-hard problems, cannot not be solved efficiently by mathematical programming techniques. Although the heuristic approaches do not guarantee to obtain the optimal solution, they can solve the VRP problem more efficiently. Under real-time information, the routing solution could be improved through exchanging demands among vehicles.
This research proposes improved petal methods to perform vehicle assignment. During the route updating process, the Swap-exchange and Tabu Search algorithms are deployed to update routes and exchange demand points between commercial vehicles. The algorithm is implemented through C++, an object-oriented language. The proposed approaches are then evaluated in a realistic traffic simulation environment. The traffic simulation-assignment model, DynaTAIWAN is applied to evaluate assigning and routing strategies in a experiment network and a true network. DynaTAIWAN is used to produce time-dependent travel cost matrix, according to different simulation scenario setups.
Numerical experiments, based on different information supply strategies, are conducted to explore the research framework for dynamic fleet management problems under real-time information.
Key words：petal method、Tabu Search algorithm、Swap Exchange、DynaTAIWAN
|
|---|
