Collaborative Research: Modeling Human Driving Behavior and Response with Applications to Intelligent Agent-Based Traffic Flow Simulation

人类在日常车辆交通情况下的决策在很大程度上决定了交通流的性质和由此产生的交通服务质量，这影响到交通系统的可靠性，安全性和效率以及城市生活的质量。 该项目提出了一种全面的、多学科的研究方法来表征和建模人类认知驾驶行为以及交通流系统中的后续响应。具体而言，驾驶员行为的动态，在个人层面上，并作为一个群体的一部分，随着时间和空间的演变将系统地研究作为一个复杂的系统。通过开发基于行为的交通状况下的人类决策模型，并将行为模型集成到计算机仿真系统中，该研究解决了交通科学中的基本问题，并有望提高对交通流现象的普遍理解以及当前交通流仿真技术的保真度和可靠性。该项目的成果将通过分析调查和模拟，并通过现场观察验证，揭示驾驶员行为的微观要素（例如反应时间和其他时滞、过度反应、风险规避和风险寻求行为）与宏观交通流现象之间的相互关系。该研究课题具有重要的经济和社会意义，在人类认知科学以及交通运营和管理中具有广泛的应用。该项目还为学生接触跨学科研究提供了一个极好的和独特的机会，并有助于建立一支多元化的科学队伍。

Human decision-making in everyday vehicular traffic situations determines in large measure the properties of the traffic stream and the resulting quality of traffic service, which affect the reliability, safety and efficiency of transportation systems and the quality of urban life. This project puts forward a comprehensive, multidisciplinary research approach to characterize and model human cognitive driving behavior and subsequent response in traffic flow systems. Specifically, the dynamics of driver behavior, taken at the individual level and as part of a group, evolving over time and space will be systematically studied as a complex system. By developing behavior-based models of human decision-making in traffic situations and integrating the behavior models in computer simulation systems, the study addresses fundamental questions in traffic science and promises to improve prevailing understanding of traffic flow phenomena as well as the fidelity and reliability of the current state of the art of traffic flow simulation. The outcome of the project will reveal the interrelationship between microscopic elements of driver behavior (e.g. reaction times and other time lags, over-reaction, risk averse and risk seeking behaviors in car following and lane changing) and macroscopic traffic flow phenomena through analytical investigation and simulation, validated by field observation. The subject of the investigation entails considerable economic and social significance, with wide application in human cognitive science, as well as in traffic operations and management. The project also provides an excellent and unique opportunity for student exposure to cross-disciplinary research and helps build a diverse scientific workforce.