Surrogate Safety Measure for Simulation-Based Conflict Study

Abstract

A surrogate measure called the aggregated crash propensity metric (ACPM) is proposed for simulation-based conflict studies to allow for relative comparisons and safety evaluations among intersection designs. A probabilistic crash propensity model is developed to determine the crash probability of simulated conflicts by introducing the distributions of reaction time and maximum braking rates. This model is able to generate the ACPM for three different crash types: crossing, rear end, and lane change. An experimental validation effort was conducted by simulating 12 intersections through the simulation package VISSIM. The surrogate safety assessment model is used to extract useful conflict data for entry into the model to estimate the ACPM. The Spearman rank tests indicate that the ACPM is able to identify the relative safety among traffic facilities and treatments. Notably, the ACPM outperforms the Highway Safety Manual procedures in rank tests. Preliminary efforts to correlate the ACPM with crashes and to use it as a crash predictor indicate that regression models are well fitted for the ACPM and predicted real crash frequency; this finding suggests that its potential is directly linked to real crashes.