Abstract: In order to improve the safety of railway freight transportation network, the importance of railway freight station was analyzed scientifically and rationally. Considering the freight stations as nodes, the goods delivery business between freight stations was abstracted as edge and transfer actions and a transportation network of goods was formulated. On this basis, the PageRank algorithm was applied to compute and obtain the importance time series sample of each station. Then, the importance of each station was abstracted as a random variable, and Gaussian mixture distribution was used to fit the importance samples to get the importance distribution function of each station. Finally, the mean value of the distribution function was taken as the measuring value of importance. A case study of Chengdu Railway Bureau Co., Ltd., was carried out to calulate. The results show that the average importance of the first-class, second-class, third-class and fourth-class stations is 4.59, 2.99, 4.24 and 2.76 respectively, which indicates that the first-class station and the third-class station are the most important station clusters in the network. Xiaozhaiba, the third-class station, is the most important node in the network and its importance value is 118.28, which is consistent with the actual data statistical analysis results. The effectiveness of the proposed method is verified.

Key words: traffic and transportation engineering; railway freight station; freight transport network; node importance; PageRank; Gaussian mixture distribution

摘要： 为了提高铁路货物运输网络安全性，科学合理分析铁路货运车站重要性，将货运车站看作节点，货运站间货物发送业务抽象为边和转移行为，构建成一个货物运输网络。在此基础上首先运用网页排序算法计算得到每个车站重要度时间序列样本，然后将每个车站重要度抽象为随机变量，接着用高斯混合分布拟合重要度样本得到各车站重要度分布函数,最后以分布函数均值作为重要度衡量数值。以中国铁路成都局集团有限公司为案例进行计算，结果表明：一、二、三、四等车站平均重要度分别为4.59、2.99、4.24和2.76，说明一等站和三等站是网络中最重要的车站集群，其中三等站小寨坝是网络中最核心节点，其重要度值为118.28，与实际数据统计分析结果一致，证明了该方法的有效性。

关键词: 交通运输工程；铁路货运车站；货物运输网络；节点重要度；网页排序；高斯混合分布