Discovering critical intersections in a road network

Abstract

A sound understanding of the importance of intersections within a road network is a key factor for the efficient and effective management and control of transport networks when traffic operations are saturated. Conventional indices (e.g. level of service, degree of saturation, capacity, or delay) reflect only the states of an intersection itself, but lack the ability to describe the influences among the intersections within a road network. This study specifically presents a novel index, the intersection rank, to evaluate the importance of an intersection within a road network; this index refers to the view of the PageRank algorithm and further considers static connection structures and dynamic influences of traffic volumes between intersections. The experiment, conducted using a simulated road network with 30 intersections, shows that the presented method has superior capability in discovering critical intersections and reveals the interrelationships between intersections. These findings can be used to discover important intersections and adjust scarce spatiotemporal resources of the road transportation system during saturated traffic operations; moreover, they will be helpful for practitioners in undertaking efficient control and management of the road network.