Title: Inference for origin-destination matrices of transport networks: a Bayesian approach
Abstract:
Information on the origin-destination (OD) matrix of a transport
network is a fundamental requirement in much transportation planning. A
relatively inexpensive method to update an OD matrix is to draw inference
about the OD matrix based on a single observation of traffic flows on a
specific set of network links, where the Bayesian approach is a natural
choice to combine the prior knowledge about the OD matrix and the current
observation of traffic flows. The existing approaches of Bayesian modeling
of OD matrices include using normal approximations to Poisson distributions
which leads to the posterior being intractable even under some simple
special cases, and/or using MCMC simulation which incurs extreme demand of
computational efforts. In this paper, through the EM algorithm, Bayesian
inference is reinvestigated for a transport network to estimate the
population means of traffic flows, reconstruct traffic flows, and predict
future traffic flows. It is shown that the resultant estimates have very
simple forms with minimal computational costs incurred.