Bayesian hierarchical modelling of traffic flow - With application to Malta's road network

Abstract

A Bayesian Hierarchical Model is presented to estimate route choice preferences between OD pairs. The methodology adopted utilizes both Origin-Destination (OD) information and traffic counts observed on some of the links in the network to estimate route choice probabilities. Route choice preferences are represented by multinomial distributions and estimated via a Markov Chain Monte Carlo (MCMC) algorithm. The proposed model takes into account measurement errors in the link counts, the uncertanties present in OD data and alternative routes choices both inside or outside the network of study. The proposed method is validated on both a synthetic example and the traffic network of Malta.