connected vehicle

emergency braking

driver behavior

assistance systems

IEEE Transactions on Vehicular Technology

Automatic emergency braking (AEB) system play an essential role in avoiding traffic collisions and minimizing impact strengths, which contributes to improving road traffic safety. However, the present research of the AEB system is mainly concentrated on avoiding the car to car and car to pedestrian collisions. China has typical mixed traffic characteristics, especially the frequent car to two-wheeler collision accidents. Research on the AEB system to prevent car to two-wheeler collisions in complex multi-traffic scenarios is of great significance to further improve the comprehensive performance of the AEB system. In this article, based on the cases of the car to two-wheeler collisions from China In-Depth Accident Study (CIDAS) database, the effects of three key parameters of the AEB system, braking deceleration, braking advance time, and detection range, on the accident rate and severity are studied by accident reconstruction and virtual experiment. Results show that the theoretical optimal detection range of the AEB system is 180°–35 m for the car to two-wheeler collision prevention. Moreover, the current feasible optimal detection range is 120°–35 m due to the limitation of environmental awareness technology. When the braking advance time is 1 s, the detection angle is 120°, and the detection distance is 35 m, the installation of the AEB system can avoid 22.3% of the car to two-wheeler collision accidents. Meanwhile, in the inevitable accidents, the average collision speed of vehicles decreases from 33.93 to 21.34 km/h, which effectively reduces the collision strengths. Findings of this study could help car manufacturers and makers of AEB hardware select the appropriate detection radar. They can also provide ideas and data support for formulating relevant laws and regulations of the AEB system, future development of AEB system, or other driver assistance systems.

Investigation on AEB Key Parameters for Improving Car to Two-Wheeler Collision Safety Using In-Depth Traffic Accident Data