Crash probability estimation via quantifying driver hazard perception

Autor: Yang Li, Kenji Kodaka, Keqiang Li, Jianqiang Wang, Yang Zheng
Rok vydání: 2018
Předmět:
Zdroj: Accident Analysis & Prevention. 116:116-125
ISSN: 0001-4575
DOI: 10.1016/j.aap.2017.05.009
Popis: Crash probability estimation is an important method to predict the potential reduction of crash probability contributed by forward collision avoidance technologies (FCATs). In this study, we propose a practical approach to estimate crash probability, which combines a field operational test and numerical simulations of a typical rear-end crash model. To consider driver hazard perception characteristics, we define a novel hazard perception measure, called as driver risk response time, by considering both time-to-collision (TTC) and driver braking response to impending collision risk in a near-crash scenario. Also, we establish a driving database under mixed Chinese traffic conditions based on a CMBS (Collision Mitigation Braking Systems)-equipped vehicle. Applying the crash probability estimation in this database, we estimate the potential decrease in crash probability owing to use of CMBS. A comparison of the results with CMBS on and off shows a 13.7% reduction of crash probability in a typical rear-end near-crash scenario with a one-second delay of driver's braking response. These results indicate that CMBS is positive in collision prevention, especially in the case of inattentive drivers or ole drivers. The proposed crash probability estimation offers a practical way for evaluating the safety benefits in the design and testing of FCATs.
Databáze: OpenAIRE