A method of road network vulnerability identification taking into account travelers' heterogeneous risk attitudes

Abstract

The analysis of road network vulnerability is a challenging and important research subject in the field of transportation reliability engineering because of the complex coupling relationships among travelers, vehicles, roads and environment. In view of the deficiency of existing researches on the description of travelers’ risk averse and bounded rational behavior characteristics, both random utility theory and random regret theory are used to describe travelers’ decision-making behaviors. Then a traffic assignment model expressed by variational inequality is constructed, in which travelers’ risk aversion and bounded rationality as well as resultant heterogeneity of travelers’ decision-making behaviors are explicitly taken into account. Subsequently, the vulnerability index of road network based on accessibility is defined, and a vulnerability identification model is built, then corresponding heuristic algorithm is also proposed. The example results show that the consequences of link closures could be misjudged and the vulnerability rankings could be misidentified, if ignoring the effects of the congestion levels of road network, travelers’ perception errors and regret aversion degrees, as well as travelers’ route choice decision criteria. Therefore, it is necessary to capture the travelers’ behavior characteristics in the process of vulnerability analysis.