Origin-destination matrix estimation with a conditionally binomial model

Abstract

A doubly stochastic, conditionally binomial model is proposed to describe volumes of vehicular origin-destination flows in regular vehicular traffic, such as morning rush hours. The statistical properties of this model are motivated by the data obtained from inductive loop traffic counts. The model parameters can be expressed as rational functions of the first and second order moments of the observed link counts. Challenges arising from the inaccuracy of moment estimates are studied. A real origin-destination traffic problem of Tampere city is solved by optimisation methods and the accuracy of the solution is examined.