交通状态划分的参数权重聚类方法研究

由于交通流量、速度、占有率或密度等参数在交通状态划分中作用不同，本文提出了基于参数权重聚类的交通状态划分方法.根据交通参数数据的相似性，应用基于加权欧氏距离的相似性度量方法构建了交通参数评价函数，并用梯度下降法极小化评价函数对交通参数权重进行求解.将交通参数权重应用于模糊C均值聚类算法（FCM），得到基于参数权重的FCM道路交通状态划分方法.应用提出的模型对选取的实际交通参数数据进行交通状态划分，并与基于欧式距离的FCM状态划分结果对比.研究结果表明，本文提出的方法提高了交通状态划分精度，更接近交通实际运行状况.

Traffic State Classification Based on Parameter Weighting and Clustering Method

Considering the different effect of each traffic parameter (traffic flow, speed, occupancy or density) on traffic state classification，the method is proposed to classify traffic state of urban traffic based on the parameter weighting. According to the similarity measurement method, a parameter evaluation function is put forward to give each parameter a parameter weighting based on weighted Euclidean distance, and minimize the function using the gradient method. After obtaining parameter weighting values, this paper uses the weighted Euclidean distance to replace the common Euclidean distance in Fuzzy C- means Clustering (FCM). Finally, we classify the traffic state using the proposed methodology, the traffic parameter data come from the road network, and comparing to the method of FCM. The results show that the proposed methodology which classifies the traffic state classification is more consistent with actual conditions.