Cluster Analysis of Probabilistic Origin-destination Demand Using Day-to-day Traffic Data

Over the past few decades, the deterministic OD demand has been used for transportation planning and management. Incidents, holidays, events, and many other factors would affect the OD demand that varies by day, and hence the deterministic OD demand fails to capture the uncertainty and stochasticity of demand. This paper proposes a cluster analysis framework for discovering patterns of probabilistic OD demand in a recurrent network. The framework adopts a distance measure specialized for transportation systems, considering the physical features of real-world networks, including travel time, network structure, correlation among links, measurement errors, etc.. An Expectation-Maximization (E-M) algorithm is proposed to obtain the estimation of mixed distribution of probabilistic OD demand. The generalized statistical traffic assignment model and the probabilistic OD estimation model are encapsulated in the proposed E-M algorithm such that the statistical features of OD demands and network conditions can be estimated within each cluster. The authors perform the cluster analysis on a small network as well as a real-world network to analyze its performance and efficiency, and to provide insights for real-world applications of clustering probabilistic demand for transportation system management.

  • Supplemental Notes:
    • This paper was sponsored by TRB committee ADB30 Standing Committee on Transportation Network Modeling.
  • Corporate Authors:

    Transportation Research Board

  • Authors:
    • Zhang, Pengji
    • Ma, Wei
    • Qian, Zhen (Sean)
  • Conference:
  • Date: 2019


  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References; Tables;
  • Pagination: 6p

Subject/Index Terms

Filing Info

  • Accession Number: 01698265
  • Record Type: Publication
  • Report/Paper Numbers: 19-05181
  • Files: TRIS, TRB, ATRI
  • Created Date: Dec 7 2018 9:50AM