Incorporating safety reliability into route choice model: Heterogeneous crash risk aversions

In this study, a route choice model which accounts for both travelers' safety concern-route safety reliability-and travel time concern is proposed. Route safety reliability (variability) is defined by the distribution of the travel crash risk cost (CRC) to represent the safety condition of travel routes. We further associate the travel safety variability due to stochastic crash occurrence with travelers' crash risk aversion route choice behaviors, and postulate that travelers acquire the variability of route travel safety based on the past experience and factor it into their route choice in the form of an effective crash risk cost (effective CRC). This effective crash cost is formed depending on travelers' requirements on safe arrivals, and thus varies with individuals and trip specific factors (e.g. purposes). Moreover, all travelers want to minimize the summing of their travel time and their effective CRC. A route-based solution algorithm is designed to solve the route choice model. Two networks including Nguyen and Dupis' network and Sioux falls network are conducted as numerical studies to illustrate the model. The results show that (1) the travelers' route choice behaviors are sensitive to the route safety performance, including the average safety condition (the mean of the CRC distribution) and safety reliability (the standard deviation of the CRC distribution); (2) the safety performance of movements at intersection would significantly influence the travelers' route choice decisions; and (3) travelers with different safety attitudes (heterogeneous crash risk aversions) would make different route choice decisions. (C) 2020 Elsevier Ltd. All rights reserved.