摘要
针对汽车主动避撞系统的安全报警要求,在比较固定车距报警、车头时距报警和驾驶员预估模型报警算法,并结合汽车制动过程运动学分析的基础上,提出了一种新的基于驾驶员模型的安全报警算法。该方法具有贴近驾驶习惯且易于工程实现的特点,该文通过仿真分析验证了其合理性。实车实验结果表明:车速为20km/h时,驾驶员模型算法优于车头时距算法,而两种驾驶员模型算法之间则相差不大;车速增加至30km/h时,驾驶员预估模型算法与车头时距算法结果接近,而该文提出的算法的报警时刻更好。
A driver model based warning algorithm was developed for vehicle collision avoidance systems based on the comparison of the fixed-distance algorithm, the time-headway algorithm, and the driver-model algorithm with the analysis of braking process. The warning algorithm simulates the driver's feelings and is easily implemented. Its effectiveness was verified by simulations and experiments. The results show that the driver-model algorithms are better than the time-headway algorithm at a velocity of 20 km/h with similar driver-model algorithms results, and that the proposed methed is the best among the time-head way algorithm and the two driver-model algorithms when the velocity reaches 30 km/h.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2004年第5期697-700,共4页
Journal of Tsinghua University(Science and Technology)
基金
国家"九五"科技攻关项目(2002BA404A21)
