Abstract: Through the analysis of urban road traffic characteristics, traffic flow parameters under the same traffic state were dispersed in a two-dimensional region. The traffic condition was divided into four states, and state transitions of traffic flow were described. According to the ambiguity characteristics of urban road traffic state, a real-time traffic condition identification method based on the fuzzy c-means clustering was presented, and the flow, speed and occupancy were taken as feature attribute of sample data. Firstly, fuzzy C-means clustering technique was used to classify the sampled historical data, and the clustering center of different traffic condition was gotten with the method, then in the test module, the real-time traffic data were used to identify which state the traffic data belong to. Finally, the traffic condition of Wenming Avenue was tested and analyzed through the actual collected data with the method. The results are same with the measured results of traffic condition, and it verified the effectiveness of this method.

Key words: traffic engineering, urban road, traffic flow, traffic congestion, traffic state identification, fuzzy C-means clustering

摘要： 根据城市道路交通流特性，针对同一交通状态下交通流参数分散在一个二维区域的现象，将交通流划分为4个状态，讨论了不同状态之间的转变情况;针对城市道路交通状态存在模糊性的特点，以流量、速度、占有率作为样本数据的特征属性，提出了基于模糊C均值聚类(FCM)的交通状态实时判别方法，该方法首先采用模糊聚类技术对历史数据进行分类，得到不同交通状态的聚类中心，然后对新观测到的交通数据所属类别进行实时判别以确定交通状态。以赣州市文明大道为实例进行分析，其结果与实测交通运行状况结果一致，验证了方法的有效性。

关键词: 交通工程, 城市道路, 交通流, 交通拥挤, 交通状态判别, 模糊C均值聚类