This thesis proposes a framework and method for dynamic origin-destination demand estimation. OD estimation is a critical component of a Dynamic Traffic Assignment system in that it determines the frequencies of drivers' trips through a network. The OD estimation method presented here allows for tunable optimization to three classes of objectives: Assigned traffic flows, deviation from historical data, and relative proportions in historical data. The method can be easily extended to make use of other sources of information such as direct measurements of OD flows from probe vehicles. The framework is extended to allow for nonnegativity and capacity constraints on the OD flows. As OD estimation is intended for use in a real-time setting, computational issues are critical, and several simplifications to increase computational efficiency are propsed and evaluted, called the Exact-Match estimator and the Large-Flow estimator. The algorithms presented are implemented as part of the DynaMIT (Dynamic Network Assignment for the Management of Information to Travelers) traffic estimation and prediction software, which incorporates models for driver route choice and traffic movement simulation.


    Access

    Download


    Export, share and cite



    Title :

    Estimation of origin-destination flows for dynamic traffic assignment


    Contributors:

    Publication date :

    2001


    Size :

    112 pages , 6618643 byte , 6618401 byte


    Remarks:

    Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2001.
    Includes bibliographical references (p. 107-112).


    Type of media :

    Theses


    Type of material :

    Electronic Resource


    Language :

    English




    Dynamic Estimation of Origin-Destination Flows front Link Traffic Counts

    Matsui, H. / Matsumoto, Y. | British Library Conference Proceedings | 1995


    Deep Learning-Based Dynamic Traffic Assignment With Incomplete Origin–Destination Data

    Fan, Wenbo / Tang, Zhenkun / Ye, Pengyao et al. | Transportation Research Record | 2022