多源数据融合的山地轨道交通站点分类研究  

Classification of Mountain Rail Transit Stations with Multi-source Data Fusion

作　　者：邹庆茹 郭雪莉 ZOU Qingru;GUO Xueli(College of Traffic and Transportation,Chongqing Jiaotong University,Chongqing 400074,China)

机构地区：[1]重庆交通大学交通运输学院,重庆400074

出　　处：《石家庄铁道大学学报(自然科学版)》2025年第1期97-103,共7页Journal of Shijiazhuang Tiedao University(Natural Science Edition)

基　　金：国家自然科学基金(52302386);中国博士后科学基金(2023M730430);重庆市教委项目(KJQN202300741)。

摘　　要：为探究山地轨道交通车站客流波动变化特征,利用自动售检票(AFC)数据、兴趣点(POI)数据、路网拓扑结构及车站周边信息,提出考虑山地特征的轨道交通车站类型辨识方法。首先,基于AFC数据和POI数据,采用肘部分析法确定合理的聚类数量,利用K-means聚类法将车站分为3大类;其次,利用路网拓扑数据和车站周边的公-轨数据进行二次聚类,最终将车站分为6个子类:综合类山地型站点、综合类非山地型站点、就业类山地型站点、就业类非山地型站点、居住类山地型站点、居住类非山地型站点。分类结果显示,综合类非山地型相对综合类山地型地势平缓,居民活动聚集性特征明显;居住类山地型与居住类非山地型相比,路网密度较小、受地势影响较大、土地开发程度低;就业类非山地型与就业类山地型相比,周边公交便捷程度高、乘客出行活跃,但停车场数量偏少。In order to explore the changing characteristics of passenger flow fluctuation in mountain rail transit stations,a method for identifying rail transit station types considering mountainous features was proposed using the automatic fare collection(AFC)data,point of interest(POI)data,road network topology,and the data of the mountain station.Firstly,based on AFC data and POI data,the reasonable cluster number was determined according to elbow analysis,and the K-means clustering method was used to divide stations into three categories;Secondly,using the road network topology data and public-rail data for secondary clustering,the stations were ultimately divided into six sub-categories:comprehensive mountain stations,comprehensive non-mountain stations,employment mountain stations,employment non-mountain stations,residential mountain stations and residential non-mountain stations.The classification results show that the comprehensive non-mountain type has a relatively flat terrain and obvious characteristics of residential activity aggregation compared to the comprehensive mountainous type;Compared with non-mountainous residential type,mountainous residential type has lower road network density,is more affected by terrain,and has lower degree of land development;Compared with employment non-mountain type,the employment mountain type has higher convenience in surrounding public transportation and active passenger travel,but the number of parking lots is relatively small.

关 键 词：车站分类 数据挖掘 城市轨道交通 多源数据 聚类 

分 类 号：U291[交通运输工程—交通运输规划与管理]