Considering perception errors in network efficiency measure: an application to bridge importance ranking in degradable transportation networks

Abstract

In this paper, a new network efficiency measure that takes into account of traveller's perception errors as well as flows, behaviours, and costs is developed for assessing link importance in congested transportation networks. The core component of the network efficiency measure is a probabilistic route choice model embedded in a network equilibrium framework, which explicitly captures the interactions between travellers’ perception errors and link cost functions to produce a stochastic user equilibrium (SUE) flow pattern. The importance measure defined as the relative network efficiency drop is developed to rank the links in the network. The network efficiency and importance measures can work for both uncongested and congested transportation networks. To show proof of concept, we use two networks to demonstrate the features and applicability of the proposed measures. In the Braess network, we examine the value of link efficiency, link importance, and link ranking in comparison with the topological measure and the user equilibrium (UE) measure. In addition, we investigate the effects capacity degradation level, demand level, and degree of perception error. In the Winnipeg network, we apply the network efficiency and importance measures to assess the bridges in the city of Winnipeg, Manitoba, Canada. We compare the SUE importance with the localised volume to capacity (V/C) ratio method, and investigate the bridge importance measures in the context of a flooding scenario to examine the effect of providing information to drivers. The results demonstrate that the SUE network efficiency and importance measures are useful in both theory and applications.