基于多源轨迹数据的城市交通状态精细划分与识别

针对基于路段的城市交通状态分析方法的不足，本文利用公交车和出租车轨迹数据提出了城市交通状态精细划分和识别方法，实现城市交通状态分析.对两种轨迹点的速度值和空间位置值分别进行归一化处理，以此为属性数据，通过迭代计算轮廓系数确定k 值完成轨迹点聚类，结合二次处理方法对类簇进行拆分和融合以划分道路交通状态；在特征级建立多源数据融合方法，实现交通状态速度值计算；以归一化后的速度值为属性数据，通过聚类将样本分为4类对应4种城市交通流状态层级.实验表明，本文方法能够实现道路交通状态精细划分，能有效地识别出道路局部位置的交通状态，进而可为城市道路交通管理提供决策支持.

Fine Division and Identification of Urban Traffic Status Based on Multi-source Trajectory Data

Considering the shortcomings of urban traffic status analysis method based on road segments, this paper proposes a fine division and identification method of urban traffic status by using bus and taxi trajectory data, realizing the urban traffic status analysis. Firstly, the velocity and spatial position values of trajectory points are normalized separately, which are used as attribute data, trajectory points are clustered by iteratively calculating contour coefficients to determine k values , and the clusters are split and merged to divide road traffic status according to the proposed cluster quadratic processing method. Next, a multi- source data fusion method is established at the feature level to calculate the traffic status speed value. Finally, the sample is divided into four categories corresponding to four urban traffic flow status levels by clustering with attribute data of the normalized velocity value. The experimental results show that the proposed method can realize the fine division of road traffic status and effectively identify the traffic status of different locations of a road, which can provide decision support for urban road traffic management.