| Summary |
Chang-Ming Wang，2006.06
Graduate Institute of Transportation Science, Tamkang University
Trip origin-destination (OD) is one of the crucial factors that affect traffic state in a highway network. Network OD data depict both the spatial and temporal distributions of corresponding trip demands, including travel direction, route choice, departure time, and trip length, etc. Therefore, it is one of the key components in transportation planning, network design, and terminal selection.
However, in practice it is not easy to collect true OD trip demands. As the development of traffic flow data collection technology, link traffic flows data are collected by vehicle detector and used to infer network OD trip demands. Therefore, an effective vehicle detector deployment plan is one of the key issues in modern traffic control and management.
The vehicle detector deployment configuration problem is essentially a long-term transportation planning issue. The purpose of the present research is to conduct route choice behavior description and model solving by traffic assignment theorem. Furthermore, we solve the user equilibrium (UE) problem by gradient projection method (GP) to obtain some additional path-based information. Since true OD trip demands are unknown, we conduct OD demand estimation and OD demand covering rate verification by the least square estimator. To summarize all of above, the present research is based on static transportation planning to construct mathematical analytical models and model solving and analysis.
To demonstrate the feasibility of the proposed models, some simplified networks including small, middle, and large network sizes were employed to conduct the case study under original trip OD demands and twice of the OD demands. The numerical results indicate that using finite quantity of vehicle detectors and deployed it strategically can obtain effective network trip OD information coverage. Besides, the number of deployed vehicle detectors increase with the scale of network size and network OD demands. The result of linearly independent deployment configuration strategies indicate that by only means of prior OD demands can find the relatively important links of all without link traffic flow measurements. Moreover, linearly independent deployment configuration strategies can obtain excellent OD demand covering rate on the cases of small and middle network, and satisfactory results for the large network cases. |
