高速公路驾驶员行为模型的建立与仿真  

MODELING AND SIMULATION DRIVER BEHAVIORS OF THE TRAFFIC ON THE HIGHWAY

作　　者：甘霖[1] 李灵犀[1] 王飞跃[1] 

机构地区：[1]中国科学院自动化研究所智能控制与系统工程中心,北京100080

出　　处：《模式识别与人工智能》2003年第3期357-362,共6页Pattern Recognition and Artificial Intelligence

摘　　要：传统的高速公路模型往往忽略了驾驶员微观个性的差别,很难满足控制、诱导等方面的精度要求。本文依据实际数据聚类分析结果将高速公路上行驶车辆的驾驶员分为五类,提出新的换道模型,并对高速公路上的自由流和拥挤流两种交通流的速度一密度关系用模糊推理进行建模,然后用实际数据进行仿真;最后对仿真结果进行了讨论,分析了模型的可行性。Traditional highway traffic models are most focused on macroscopic traffic, ignoring the microdifference a-mong drivers, which made such kind of models diffcult to meet the precision requirements in traffic control and cruising. In this paper, first we classify the drivers on the highway into five types according to the result of cluster analysis. And then we propose the flow-density relations under congested flow regime and noncongested flow regime respectively on the highway by fuzzy reasoning. Finally, we simulate our behavior models by real data and give our results evaluation.

关 键 词：高速公路 驾驶员 驾驶员行为模型 仿真 模糊推理 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]