Combination of vehicle routing models and dynamic traffic simulation for city logistics applications
The urban network is a highly dynamic system. Thus, a modern and efficient fleet management in urban areas should account for dynamics of traffic conditions, variability in travel times, changes in demand and fleet availability. All these factors significantly affect the distribution of goods and th...
| Autor: | |
|---|---|
| Tipo de documento: | tese |
| Data de publicação: | 2012 |
| País: | España |
| Recursos: | Universitat Politècnica de Catalunya (UPC) |
| Repositório: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglês |
| OAI Identifier: | oai:upcommons.upc.edu:2117/95002 |
| Acesso em linha: | https://hdl.handle.net/2117/95002 https://dx.doi.org/10.5821/dissertation-2117-95002 |
| Access Level: | Acceso aberto |
| Palavra-chave: | Enginyeria del trànsit -- Mètodes de simulació Àrees temàtiques de la UPC::Matemàtiques i estadística |
| Resumo: | The urban network is a highly dynamic system. Thus, a modern and efficient fleet management in urban areas should account for dynamics of traffic conditions, variability in travel times, changes in demand and fleet availability. All these factors significantly affect the distribution of goods and the provision of services. As a consequence, the freight operations optimizing approaches should be based on the time-dependent travel time estimates rather than on the average static values commonly employed as input data. Proficient dynamic fleet management decisions need to take into consideration all the factors conditioning the addressed problem. Hence, the customers’ requests and service conditions (demands, time windows, etc.), operational conditions of the employed fleet (vehicles’ availability, status, positions, current occupancy of the carriage space, etc.) and traffic conditions need to be reckoned with. This information can be provided in the real-time fashion and at an affordable price by the ICT applications and tools such as: ATIS, GPS, GPRS or other. Instead of relying exclusively on the experience of a dispatcher, it is effective to base the freight management decisions on information provided by a professional Decision Support System facilitating the consideration of all the factors conditioning the addressed issue. The objective of the present thesis is to propose, implement and computationally test, on the basis of a simulation, the architecture of a Decision Support System for real-time freight management able of accounting for all the dynamic factors mentioned above. Its design is based on integration of a selected pickup and delivery vehicle routing model and dynamic traffic simulation models, whose purpose is to carefully emulate the evolving traffic conditions. The optimal dynamic routing and scheduling of a vehicle fleet is obtained due to dynamic modifications of the current routing and scheduling plan on the basis of the recently revealed information conditioning the addressed problem. The obtained results of computational experiments show that the performance of freight fleet strongly depends on the traffic information used to create the routing and scheduling plan. Due to the special character of transport operations performed in metropolitan areas it is possible that not all the scheduled customers would be served, although a feasible initial solution is created on the basis of the historical travel times’ data. Hence, there is a need to implement real-time rerouting strategies allowing modifications of the original routes in order to feasibly fulfil the routing tasks. In addition, the tighter the constraints specifying the customers’ time windows and the higher the number of dynamically appearing requests, the higher the cost of the performance and the level of utilization of the freight fleet. |
|---|