Chair of Transportation Systems Engineering, TUM School of Engineering and Design, Technical University of Munich, Arcisstrasse 21, 80333 Munich, Germany.
Chair of Transportation Systems Engineering, TUM School of Engineering and Design, Technical University of Munich, Arcisstrasse 21, 80333 Munich, Germany.
Accid Anal Prev. 2024 Aug;203:107604. doi: 10.1016/j.aap.2024.107604. Epub 2024 May 10.
The interactions of motorised vehicles with pedestrians have always been a concern in traffic safety. The major threat to pedestrians comes from the high level of interactions imposed in uncontrolled traffic environments, where road users have to compete over the right of way. In the absence of traffic management and control systems in such traffic environments, road users have to negotiate the right of way while avoiding conflict. Furthermore, the high level of movement freedom and agility of pedestrians, as one of the interactive parties, can lead to exposing unpredictable behaviour on the road. Traffic interactions in uncontrolled mixed traffic environments will become more challenging by fully/partially automated driving systems' deployment, where the intentions and decisions of interacting agents must be predicted/detected to avoid conflict and improve traffic safety and efficiency. This study aims to formulate a game-theoretic approach to model pedestrian interactions with passenger cars and light vehicles (two-wheel and three-wheel vehicles) in uncontrolled traffic settings. The proposed models employ the most influencing factors in the road user's decision and choice of strategy to predict their movements and conflict resolution strategies in traffic interactions. The models are applied to two data sets of video recordings collected in a shared space in Hamburg and a mid-block crossing area in Surat, India, including the interactions of pedestrians with passenger cars and light vehicles, respectively. The models are calibrated using the identified conflicts between users and their conflict resolution strategies in the data sets. The proposed models indicate satisfactory performances considering the stochastic behaviour of road users - particularly in the mid-block crossing area in India - and have the potential to be used as a behavioural model for automated driving systems.
机动车与行人的相互作用一直是交通安全关注的焦点。对行人的主要威胁来自于在不受控制的交通环境中交互作用的高度，在这种交通环境中，道路使用者必须争夺先行权。在这种交通环境中没有交通管理和控制系统，道路使用者必须在避免冲突的同时协商先行权。此外，作为交互方之一的行人具有较高的运动自由度和灵活性，可能导致其在道路上表现出不可预测的行为。在完全/部分自动驾驶系统部署的情况下，不受控制的混合交通环境中的交通交互将变得更加具有挑战性，因为必须预测/检测交互代理的意图和决策，以避免冲突并提高交通安全和效率。本研究旨在制定一种博弈论方法来模拟行人与乘用车和轻型车辆（两轮和三轮车辆）在不受控制的交通环境中的相互作用。所提出的模型采用了影响道路使用者决策和策略选择的最主要因素，以预测他们在交通交互中的运动和冲突解决策略。该模型应用于在德国汉堡的共享空间和印度苏拉特的中间街区交叉口收集的两个视频记录数据集，分别包括行人与乘用车和轻型车辆的相互作用。该模型使用数据集中用户之间的已识别冲突及其冲突解决策略进行校准。所提出的模型考虑到了道路使用者的随机行为，表现出令人满意的性能-特别是在印度的中间街区交叉口-并且具有作为自动驾驶系统行为模型的潜力。
