A Discrimination Method for Traffic Conflict Severity on Slow-moving Shared Paths

To analyze the traffic safety on slow-moving shared paths, this study proposes a method for discriminating the severity of conflicts in the scenario, using traffic conflicts as an alternative safety measure. Initially, utilizing video-based automatic recognition technology to extract conflict trajectories. Subsequently, indicators for conflict severity are computed and binary logit models are constructed to analyze the effectiveness of the indicators based on evaluators' assessments of conflict severity. Finally, an aggregate conflict index (ACI) that combines effective indicators is proposed to measure the severity of conflicts. The example results indicate that on slow-moving shared paths, the indicators of the minimum time to collision (TTCmin) and yaw rate ratio (YRR) can effectively discriminate conflict severity, while the deceleration rate to avoid collision (DRAC) indicator is ineffective. Considering multiple effective indicators yields better discrimination effects than considering a single effective indicator alone. The ACI combined by TTCmin and YRR can effectively discriminate conflict severity and demonstrate better discrimination effectiveness.