城市轨道交通站点分类的聚类方法研究

在城市轨道交通相关研究中,需要对站点进行划分,讨论不同站点的差异性。采用聚类分析的方法,选取站点自身特点和站点环境特征等相关的11个因素作为聚类分析的初始变量,对变量进行量化和标准化。对标准化的变量进行因子分析,从变量中提取隐藏的三个公共因子:步行环境因子、站点规模因子、站点接驳因子,达到突出特点和降低变量维度的效果。采用K-均值法,根据提取的公共因子进行聚类,最终将成都地铁1号线现运营16个站点划分为五大类。

Classifying Urban Rail Transit Stations Using Cluster Analysis

The related researches on city rail transit require the classification of stations and the discussion of the differences of different stations. This paper uses cluster analysis to classify stations by selecting 11 initial variables relevant to stations characteristics and environment conditions to quantify and standardize the variables. Then,the standardized factors are analyzed and the pedestrian environment factor,station size factor and feeder factor are extracted to highlight the characteristics and reduce dimension of the variables. Finally,the 16 stations of Chengdu metro line 1 are divided into five clusters with k-mean algorithm on the basis of the extracted common factors.