融入驾驶人感知的交通流建模方法研究

项目名称： 融入驾驶人感知的交通流建模方法研究

项目编号： No.71471014

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 管理科学

项目作者： 关伟

作者单位： 北京交通大学

项目金额： 60万元

中文摘要： 本课题希望借鉴人类脑科学和现代交通流理论等相关领域研究的新成果，利用神经信息学这一门新兴的边缘学科的研究方法和手段，与现代交通流理论的研究范畴相交叉结合，探索将驾驶人作为复杂交通流系统演化过程的一个变量因素，纳入其理论研究框架中的新方法。课题将利用驾驶人心理生理特征分析设备系统，以及各种交通检测设备提供的实际交通流数据，通过设计可控实验，获取不同类型驾驶员在各种交通环境条件下的原始感知数据，探讨驾驶员感知过程的可测性。通过找到一种合适的驾驶员感知过程广义度量方法，将驾驶员在驾驶过程中的动态感知行为有机融入交通流微观和宏观模型之中，从而构建含有驾驶人行为的新交通流模型，并借以体现交通流内在的 自驱动粒子系统特性。课题试图从一个新的角度揭示交通流非线性现象和相变产生的机理，解释目前发现的各种交通现象背后的深层次原因，拓展交通流理论研究的视野和思路，丰富交通流理论研究的内容。

中文关键词： 复杂系统；交通流理论；驾驶员感知；交通行为；神经信息学

英文摘要： Based on the newest achievements of human brain science and modern traffic flow theory, this project hopes to introduce drivers' perception as a variable factor in the evolution of complex traffic flow systems, so as to explore a novel approach which will expand classical framework of the traffic flow theoretical studies. The relative approaches and tools in the new frontier Neuroinformatics and modern traffic flow theory will be integrated into this project. Firstly, utilizing the devices of analyzing drivers' mental and physical characteristics as well as traffic flow data collected by using various traffic detection sensors, the project collects raw perception data of various drivers under various traffic conditions, and explores the measurability of drivers' perception through designing controllable experiments. Secondly, by proposing a generalized method of properly measuring drivers' perception, the project incorporates drivers' dynamic perception into microscopic and macroscopic traffic flow models, and proposes new traffic flow models involving drivers' behavior in order to reveal the system characteristics of self-driven particle in the traffic flow. The project intends to unveil the mechanism of the nonlinear phenomena and phase transitions in the traffic flow, to explain the deep reasons behind various traffic phenomena, and to expand and enrich relevant investigations in traffic flow theory.

英文关键词： complex systems;traffic flow theory;driver perception;traffic behavior;neuroinformatics