Synthetic populations : a tool for estimatig travel demand.

Travel demand estimation is most often based on census or mobility surveys. However, both these data sources have drawbacks, especially if one is to consider spatially disaggregate models. Census data reaches its limit when considering fine spatial desegregation, as cross-tables at municipality level are too close to breaking the legally mandatory data anonymity. Data gathered from mobility surveys, on the other hand, is often not significant at the level of specific origins and destinations. In this paper, we consider a methodology based on constructing synthetic populations, for which anonymity and spatial desegregation are not problematic. We first describe how synthetic population can be built when local samples are poor and data consistency imperfect. In this situation, the known Iterative Proportional Fitting (lPR) becomes increasingly inadequate, which is why we propose an alternative least-squares formulation. This formulation is illustrated in building a synthetic population for Belgium at the municipality level. We conclude the paper by sketching how synthetic populations can also be coupled with demographic models to produce long-terrn forecasts of spatially desegregated mobility indicators. For the covering abstract of the conference see ITRD Abstract n°E218203.

  • Authors:
    • TOINT, P
  • Publication Date: 2005


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01143994
  • Record Type: Publication
  • Source Agency: TRL
  • Files: ITRD
  • Created Date: Nov 16 2009 12:19PM