基于 K-means 的北京地铁路网重要度聚类分析

以图论为基础，以北京地铁为研究对象，结合地铁运营客流时空分布的特点， 构建北京地铁有向加权路网模型；采用 K-means 聚类分析方法，根据地铁路网中车站和区 间的两个基本的物理拓扑属性（度、介数），以及客运量对其进行分类，确定关键车站和区 间.其中，度反映的是节点的局部聚集能力，介数反映的是节点和边对全局的影响能力，而 客运量则反映了不同时间段节点和边在运输中的重要性.实证分析表明，该方法可以从系 统网络的角度动态辨识系统中的关键车站和区间.

K-means Clustering Analysis of Key Nodes and Edges in Beijing Subway Network

This paper modeled a subway system as a directed and weighted network with consideration of the temporal and spatial distribution of passengers in the subway system. Based on the K-means clustering, stations (nodes) and intervals (edges) in a subway network were grouped by three metrics: two basic topologi- cal properties (degree and betweenness), and their roles in transporting people (passenger volume). Degree re- flects the nodes’local accumulation ability; betweenness reflects a node or edge’s the impact on the global network topology, and passenger volume reflects a node or edge’s importance in transport people at different times. Taking the Beijing Subway network as a case study, the paper tested the effectiveness of the proposed approach. The results suggested that the method could identify key nodes and edges and provide dynamic de- cision support for subway network operators.