A Utility Maximization Model of Pedestrian and Driver Interactions
A Utility Maximization Model of Pedestrian and Driver Interactions
复制标题
DOI：
10.1109/access.2022.3213363
复制
发表时间：
2021-10
期刊：
影响因子：
3.9
通讯作者：
Yi-Shin Lin;Aravinda Ramakrishnan Srinivasan;M. Leonetti;J. Billington;G. Markkula
中科院分区：
文献类型：
--
作者：
Yi-Shin Lin;Aravinda Ramakrishnan Srinivasan;M. Leonetti;J. Billington;G. Markkula
许多模型都考虑了道路使用者的交通流量，但很少有模型考虑到局部交互的细节以及它们如何恶化为安全关键情况。在现有的人类感觉运动控制模型的基础上，我们开发了一个基于代理的建模框架，应用效用最大化、运动原语和间歇性动作决策的原理来解释道路使用者之间交互行为的细节。该框架将三个原则与决策理论联系起来，并进行测试以确定这种方法是否可以重现以下四种现象。首先，当两个行人在交叉路径上行驶时，他们的交互对初始运动学不对称性很敏感，其次，基于不对称性，两个行人通过调整他们的行为来快速解决碰撞冲突。第三，当行人在面对驶近的汽车时过马路时，行人会根据汽车到达的时间来调整自己的过马行为；第四，行人或汽车驾驶员可能会向对方让行以解决冲突。我们证明这些现象是从我们的建模框架中自然出现的。我们相信所提出的行为模型和以现象为中心的分析方法为检查道路用户交互提供了有前途的工具。最后，我们讨论了如何将该模型推广到安全关键情况，并包含影响道路使用者交互的其他变量。
Many models account for the traffic flow of road users, but few take the details of local interactions into consideration and how they could deteriorate into safety-critical situations. Building on an existing model of human sensorimotor control, we develop an agent-based modeling framework applying the principles of utility maximization, motor primitives, and intermittent action decisions to account for the details of interactive behaviors among road users. The framework connects the three principles to the decision theory and is tested to determine whether such an approach can reproduce the following four phenomena. Firstly, when two pedestrians travel on crossing paths, their interaction is sensitive to initial kinematic asymmetries, and secondly, based on the asymmetries, the two pedestrians rapidly resolve collision conflict by adapting their behaviors. Thirdly, when a pedestrian crosses a road while facing an approaching car, the pedestrian adapts his or her crossing behavior according to the time-to-arrival of the car, and fourthly, either the pedestrian or the driver of the car may yield to the other to resolve their conflict. We show that these phenomena emerge naturally from our modeling framework. We believe the proposed behavior model and phenomenon-centered approach of analysis offer promising tools to examine road user interactions. We conclude with a discussion on how the model can be generalized to safety-critical situations and to include other variables affecting road-user interactions.