摘要
针对传统通勤特征测算中存在的不足,提出一种基于互联网时空大数据的通勤特征挖掘技术框架。基于互联网定位、地图数据,利用机器学习算法挖掘常驻点、提取通勤OD,基于通勤OD进一步挖掘通勤距离、通勤时间以及通勤方式,并将上述通勤特征数据应用于全国主要城市通勤监测报告和国土空间规划等方面。使用多源时空大数据对通勤监测指标和结果进行校验,结果表明基于互联网位置数据的通勤特征与抽样调查获得的通勤特征具有一致性,且能够以大样本、低成本、高空间精度提供高频更新的通勤监测指标,是对传统方法的有效补充和强化。
To overcome the shortcomings of the traditional commuting travel analyzing method,this paper develops a technical framework based on the big spatial-temporal travel data on Internet.Using Internet location and map data,the developed learning algorithms can estimate commuters'residence location and workplace as well as commuting distance,time,and travel mode selection.The commuting information has been used in commuting monitoring,national land use planning and other aspects in major cities across China.Comparing commuting travel data from various sources containing spatialtemporal information verifies that the characteristics of commuting travel retrieved and estimated from the Internet are consistent with the conventional sample surveys'results.Furthermore,the huge data size,low cost and high accuracy makes the real-time Internet commuting data the best estimate for monitoring the commuting travel and a great supplement to traditional data methods.
作者
阚长城
闫浩强
项雯怡
万涛
付凌峰
Kan Changcheng;Yan Haoqiang;Xiang Wenyi;Wan Tao;Fu Lingfeng(Baidu.com Times Technology(Beijing)Co.,Ltd.,Beijing 100085,China;Tianjin Urban Planning&Design Institute,Tianjin 300201,China;China Academy of Urban Planning&Design,Beijing 100037,China)
出处
《城市交通》
2020年第5期61-67,共7页
Urban Transport of China
关键词
通勤监测
通勤OD
通勤时间
时空大数据
commuting monitoring
commuting OD
commuting time
big spatial-temporal data
