二分K-FCM结合算法在交通运行状态判别中的应用

Application of Fuzzy C-means( FCM) with Bisecting K-means Combined Algorithm in Traffic State Identification

正确判别交通运行状态是交通运营管理的理论依据。以高速公路交通状态判别为研究对象,综合考虑交通流三参数(流量、速度、占有率)的基础上,应用模糊C均值(FCM)与二分K均值结合算法对交通运行状态进行判别。首先,对交通数据集分布特征及交通运行状态特征进行分析,确定以V05~V85为最小欧氏距离判别的数据范围。其次,为解决算法收敛较慢及任意初始化质心对聚类结果的不良影响,对传统模糊C均值聚类算法进行了改进,将运行二分K均值算法的聚类结果矩阵作为FCM的初始聚类中心。经检验,改进的FCM可以有效减少算法迭代次数,得到的目标路段交通状态判别矩阵能较精准地划分高速公路不同的交通状态。

Correctly identifying real-time traffic operational condition is the basis of traffic operation and management.The paper proposed a new algorithm model which is developed from the fuzzy c-means algorithm(FCM)and bisecting K-means algorithm to identify the real-time traffic operational condition based on the integrated consideration of three traffic flow parameters.First,the paper adopted V05~V85 traffic data as the minimum Euclidean distance data scope through analyzing characteristics of the traffic data and status.Then,the clustering result matrix of bisecting K-means algorithm is used as the initial clustering center of FCM to improve the traditional fuzzy C-mean clustering algorithm in order to solve the problem of slow algorithm convergence and the adverse influence of the initial centroid on the clustering results,and it turned out to be a good solution to reduce the number of iterations.It is verified that the traffic status discrimination matrix of the target road can distinguish the different traffic states of the expressway accurately.