基于K-Means聚类算法的城市轨道交通站点分类及客流特征分析

文章立足于沈阳市轨道交通站点800 m范围内的人口分布、开发强度、公交接驳、路网长度分布、站点位置属性、站点客流数据等六大维度数据,采用K-Means聚类算法将城市轨道交通站点划分为居住型、商业商务型、综合开发型、产业型、交通枢纽型五大类;基于站点聚类成果,叠合多样化数据分析总结各类站点客流的普适性规律,可为后续站点周边基础设施完善、站点客流预测、车站运营组织方案做出指导。

Classifi cation of urban rail transit stations and points and analysis of passenger fl ow characteristics based on K-Means clustering algorithm

This paper is based on Shenyang rail transit station 800 m radius vicinity based on the six dimensions of population distribution,development intensity,public transport connection,network length and distribution,station location features and passenger flow data.The urban rail transit stations are classified into five categories:residential type,commercial type,comprehensive development type,industrial type and transportation hub type by using K-Means clustering algorithm,based on the results of station clustering,the total data is analyzed by taking into consideration of the diversifi ed data.The conclusion of the universal pattern of passenger flow at various stations provide guidance for the improvement of infrastructure around the subsequent stations,passenger flow forecast at stations,and station operation organization scheme.