Safety assessment of driving behavior in multi-lane traffic for autonomous vehicles
Safety assessment of driving behavior in multi-lane traffic for autonomous vehicles
复制标题
自动驾驶汽车多车道交通中驾驶行为的安全评估
DOI：
--
复制
发表时间：
2009
期刊：
影响因子：
--
通讯作者：
M. Buss
中科院分区：
文献类型：
--
作者：
M. Althoff;O. Stursberg;M. Buss
提出了一种自动评估自动车辆计划轨迹的安全方法。由于道路交通的不安全性，所提出的安全评估是在概率设置中进行的，其中基于动态模型计算交通参与者未来可能位置的概率分布。底层算法必须高效，因为它们需要在车辆运行时提供当前规划路径的安全级别。为了实现必要的计算效率，交通参与者的动态被抽象为马尔可夫链。该方法考虑了车辆动力学、交通参与者之间的交互以及多车道交通中的车道变换。
An approach for the safety assessment of planned trajectories of autonomous vehicles is presented. Due to the unsafe nature of road traffic, the proposed safety assessment is performed in a probabilistic setting, where the probability distributions of possible future positions of traffic participants are computed based on dynamic models. The underlying algorithms have to be efficient as they need to provide a safety level of the currently planned path while the vehicle is in operation. In order to achieve the necessary efficiency of computation, the dynamics of traffic participants is abstracted into Markov-chains. The approach considers vehicle dynamics, interaction between traffic participants, and lane changes in multi-lane traffic.