Road-safety recognition and network equilibrium with perceived route-choice sets

Safety is regarded as the second basic need in Maslow's hierarchy of needs(1943), and safety recognition and circumvention behaviour in the route-choice decision-making process should therefore be accommodated in network-traffic equilibrium analysis frameworks. This paper proposes a framework by which crash frequency, forecasted using the safety-analysis method or compiled from historical data for intersections, is used to measure the safety consciousness of drivers. Drivers are then classified into different groups according to their acceptable-risk thresholds, and each group has its own route-choice set. Decision behaviour whereby drivers are willing to bear additional costs in order to circumvent travel risk is incorporated into the variational inequality model based on the user equilibrium in the perceived route-choice set(UE-PRCS), which is an extension of Wardrop's first principle. The Frank–Wolfe algorithm, based on the convex combination method, is employed to obtain the solution. A small road network is used as a case study to illustrate the proposed framework, incorporating risk recognition and circumvention behaviour under different combinations of traffic demand and risk-sensitivity group ratio. The results show that the standard user equilibrium is a special case of the UE-PRCS, but that the UE traffic state is more common than the UE-PRCS under different parameters.