Dynamic multi-region macroscopic fundamental diagram stochastic user equilibrium: model framework and parameter estimation in a real-life large-scale case study

Abstract

Multi-region Macroscopic Fundamental Diagram (MFD) traffic equilibrium models have been developed as a more easily calibratable, maintainable, and computationally efficient alternative to traditional link-network traffic assignment models with full disaggregate network representation. In this extended abstract, we discuss the framework of a new dynamic multi-region MFD stochastic user equilibrium model that can base regional path choice on region travel times actually experienced. The model produces continuous outputs, facilitating the development of a maximum likelihood estimation procedure for rigorous statistical estimation of behavioural parameters of underlying regional path choice models. This estimation procedure is operationalised in a first-of-its-kind real-life large-scale multi-region MFD system, with underlying regions and directional superimposing motorway regions. An enormous dataset of GPS records is utilised to calibrate MFD functions and estimate models. Results provide empirical evidence supporting the hypothesis that regional path choices are more realistically based on experienced region travel times rather than instantaneous travel times.