Real-Time Crash Prediction Model for Application to Crash Prevention in Freeway Traffic

TLDR

Crash likelihood is strongly influenced by short‑term traffic turbulence, necessitating real‑time estimation of crash potential, yet prior models relied on unverified assumptions. The study aims to develop evidence‑based methods for selecting crash precursors and to evaluate the performance of the revised real‑time crash prediction model. A probabilistic real‑time model linking crash potential to traffic flow characteristics was constructed and its performance was evaluated using experimental data. The revised model objectively identifies crash precursors, removing the dependence on analysts’ subjective categorization present in earlier models.

Abstract

The likelihood of a crash or crash potential is significantly affected by the short-term turbulence of traffic flow. For this reason, crash potential must be estimated on a real-time basis by monitoring the current traffic condition. In this regard, a probabilistic real-time crash prediction model relating crash potential to various traffic flow characteristics that lead to crash occurrence, or “crash precursors,” was developed. In the development of the previous model, however, several assumptions were made that had not been clearly verified from either theoretical or empirical perspectives. Therefore, the objectives of the present study were to ( a) suggest the rational methods by which the crash precursors included in the model can be determined on the basis of experimental results and ( b) test the performance of the modified crash prediction model. The study found that crash precursors can be determined in an objective manner, eliminating a characteristic of the previous model, in which the model results were dependent on analysts’ subjective categorization of crash precursors.