摘要： 采用固定元胞单元尺寸建立自行车流元胞自动机(CA)模型会产生仿真偏差。为此，提出一种基于现场视频检测数据的元胞尺寸标定方法。通过统计分析车头间距、横向间距与车速的关系，设计一个动态修正元胞长度的函数。对比修正前后模型演化获得的宏观三参数值，可以发现，修正函数能较好地描述跟车距离与车速的动态关系，克服模型因采用定长度元胞单元而造成的面积密度与流率偏差。

关键词: 自行车流, 宏观特性, 元胞自动机, K-means聚类, 元胞尺寸标定, 动态修正函数

Abstract: Aiming at correcting the deviation by simulating a bicycle Cellular Automata(CA) model with fix cellular size, a calibration process of the cellular size based on field video detecting data is elaborated here. And by analyzing the relationship of headspace, horizontal space with speed, it puts forward a function of dynamical amending cellular length. Through the comparison of the three macro parameters obtained from the model by corrected or not, it shows that this method can effectively express the relationship following distance with speed, overcome the inappropriate area density and flow rates value with the fixed cellular length.

Key words: bicycle flow, macroscopic property, Cellular Automata(CA), K-means clustering, cellular size calibration, dynamical amending function

中图分类号: 

• U491