基于紧急制动行为预测的汽车智能制动灯研究

Research on vehicle intelligent brake light based on emergency braking behavior prediction

为预防车辆紧急制动造成的追尾碰撞事故,分析跟驰行驶中相邻车辆驾驶人的制动行为过程,提出基于紧急制动行为预测的防追尾碰撞方法。以实际道路采集的驾驶行为数据为基础,定量分析紧急制动、常规制动、加速换挡等行为的加速、制动、离合器踏板的位移变化过程,建立基于加速踏板位移变化率及驾驶操纵动作时序过程的紧急制动预测模型,设计开发汽车智能制动灯。分别在虚拟驾驶平台和实际道路上,对智能制动灯进行测试试验。结果表明,虚拟驾驶试验情况下,使用智能制动灯的车辆被追尾的发生率降低了30%;实际道路测试情况下,制动灯能够有效预测驾驶人的紧急制动行为,先于常规制动灯0.1-0.3 s点亮,提醒后车驾驶人注意安全。

To prevent rear-end collision caused by emergency braking,drivers＇ braking behaviors in the car following situations were analyzed and an anti-rear-end collision method was worked out. Displacements of the gas pedal,brake pedal and clutch pedal in emergency braking,conventional braking and accelerating behaviors were analyzed quantitatively based on the driving behavior data collected in real traffic environments. A model was built for emergency braking prediction based on both rate of the accelerator pedal＇s displacement change and the driver＇s maneuvering process. An intelligent brake light based on emergency braking prediction was devolved. Experiments were conducted on a virtual driving platform and in the real traffic environment to verify the intelligent brake light. The virtual driving results indicate that the intelligent brake light could reduce rear-end collision by 30%. The road test results also indicate that the intelligent brake light could predict drivers＇ emergency braking behavior and be activated about 0. 1-0. 3 s in advance compared with the conventional brake light,thus the following driver could respond earlier.