Prediction of rear-end conflict frequency using multiple-location traffic parameters

Traffic conflicts are heavily correlated with traffic collisions and may provide insightful information on the failure mechanism factors that contribute more towards a collision. Although proactive management systems have been supported in research community, autonomous vehicles (AVs) soon to become reality, analyses concentrated very specific environments using aggregated data. This study aims at investigating –for first time- rear-end conflict frequency an urban network level vehicle-to-vehicle interactions correlating corresponding state. The Time-To-Collision (TTC) Deceleration Rate Avoid Crash (DRAC) metrics utilized estimate current situation, as well scenarios including AV characteristics. Three critical points defined, according TTC DRAC thresholds. After extracting conflicts, data fitted into Zero-inflated also traditional Negative Binomial models, quasi-Poisson while controlling for endogeneity, order investigate contributory of frequency. Results demonstrate counts significantly higher congested high variations speed increase conflicts. Nevertheless, comparison simulated use surrogate safety indicators could insight relationship between state near future.