摘要
过街行人运动不确定性是影响人-车碰撞和行人伤害预测准确性与可靠性的重要因素。本文中采用1阶马尔科夫模型描述行人过街运动,利用无味变换,高效地估计人-车碰撞概率和碰撞时间与速度的概率密度,从而预测人-车冲突中行人的伤害概率。在运动仿真的基础上,构建了无味变换点集。以蒙特卡罗模拟为参考,验证所提出方法预测行人伤害概率的准确性。采用仿真,预测在人-车冲突场景下,采取不同避撞操作的行人伤害概率。结果表明,相比于确定性方法,本文中所提出的方法能更加真实地反映在不同避撞操作下对行人的伤害,为避撞决策与控制提供有效的目标量化手段。
The uncertainty of pedestrian movement during road crossing is an important factor affecting the correctness and robustness of the prediction of vehicle-pedestrian collision and pedestrian injuries.In this paper,the first order Markov model is employed to describe the road crossing movement of pedestrian and unscented transformation(UT)is used to efficiently estimate the probability of vehicle-pedestrian collision and the probability density of collision time and speed,and hence to predict the injury probability of pedestrian in collision.Based on kinematical simulation,the-set of UT is constructed.With Monte Carlo simulation as reference,the correctness of pedestrian injury probability predicted by the method proposed is verified.The pedestrian injury probability in vehicle-pedestrian collision with different vehicle maneuvers adopted to avoid collision is predicted buy simulation.The results show that compared with the determinate method,the method proposed in this paper can more truly reflect the pedestrian injury with different vehicle maneuvers,providing an effective objective quantization means for the decision making and control of collision avoidance.
作者
黄智
何颖
刘象祎
宋晓琳
Huang Zhi;He Ying;Liu Xiangyi;Song Xiaolin(Hunan University,State Key Laboratory of Advanced Design and Manufacturing for Vehicle Body,Changsha 410082)
出处
《汽车工程》
EI
CSCD
北大核心
2018年第1期28-33,共6页
Automotive Engineering
基金
国家自然科学基金(51575169)资助.
关键词
行人
碰撞
伤害
概率
无味变换
pedestrian
collision
injury
probability
unscented transformation
