单向非机动道路内混合自行车交通状态判别

Traffic State Recognition of Mixed Bicycle Flow in One-way Non-motorized Road

为准确判别单向共享非机动道路内混合自行车交通流的状态,以流量和密度为表征指标,将模糊C均值聚类方法应用于状态判别,并采用统计回归方法分析道路特征参数、车辆特征参数和骑行者特征参数对交通流状态聚类中心的影响。结合杭州实测数据,将混合自行车交通流状态划分为畅通、稳定通行和拥挤3种状态,分别建立各状态下聚类中心流量和密度与车道宽度、电动自行车占比、男性骑行者占比3个特征参数之间的回归关系模型。结果表明,在稳定通行和拥挤状态下,聚类中心的流量都与车道宽度呈线性相关,稳定通行状态下的密度与电动自行车占比呈线性相关。

In order to accurately distinguish all kinds of states of mixed bicycle traffic flow in one-way shared bicycle path,the fuzzy C-means clustering method was applied with the traffic flow and density as the characterization indexes.The statistical regression method was used to analyze the influence of road characteristic parameters,vehicle characteristic parameters and pedestrian characteristic parameters on the traffic flow state clustering center.The mixed bicycle traffic flow state was divided into three substates including smooth,stable and crowded based on the field data in Hangzhou.The regression relationship between the flow or density of the clustering center and the three characteristic parameters which including lane width,the proportion of electric bicycle and the proportion of male cyclist were established,respectively.The results show that the flow of clustering center is linear regression with the lane width in the state of stable and crowded.The density of clustering center is linear regression with the proportion of E-bike in the state of stable.