海口市网约车乘客出行时空模式挖掘  

Excavation of the spatio-temporal pattern of passenger travel in Haikou city

作　　者：周梦杰 白紫月 高兴 王晓密 赵琛琛 ZHOU Mengjie;BAI Ziyue;GAO Xing;WANG Xiaomi;ZHAO Chenchen(College of Resources and Environmental Science,Hunan Normal University,Changsha 410006,China;Hunan Key Laboratory of Geospatial Big Data Mining and Application,Changsha 410006,China)

机构地区：[1]湖南师范大学资源与环境科学学院,长沙410006 [2]地理空间大数据挖掘与应用湖南省重点实验室,长沙410006

出　　处：《测绘科学》2021年第10期177-184,218,共9页Science of Surveying and Mapping

基　　金：国家自然科学基金项目(41901314);数字制图与国土信息应用工程自然资源部重点实验室开放研究基金资助项目(ZRZYBWD201902);湖南省教育厅科学研究项目(18C0049);湖南省大学生创新创业训练计划项目(S202010542072)。

摘　　要：针对利用起讫点数据研究出行特征导致时间与空间维度割裂的问题,该文提出了一种出行模式分析方法。基于海口市滴滴出行2017年5—10月的1200余万个订单数据,利用以鲁棒局部加权回归作为平滑方法的时间序列分解方法(STL)开展时间趋势分析。基于整体趋势,将海口市划分为1 km×1 km的网格,引入经验正交函数分解法(EOF),将海口市居民的出行时间序列分解为空间模态和时间系数两部分,以探讨居民出行规律。结果显示,在研究时段内,研究区内的出行量整体呈上升趋势;出行模式具有很强的周期性;居民出行的时空模式在工作日与周末存在差异。该文深入分析居民的出行模式,可为优化交通资源配置等提供科学依据。When analyzing the spatiotemporal patterns of travel based on origin and destination data,spatial and temporal dimensions are usually separated,to solve this problem,this study proposes a travel mode analysis method.Based on the data of more than 12 million orders of Didi Travel in Haikou from May to October 2017,the seasonal and trend decomposition using loess(STL)method is used to conduct time trend analysis.Based on the overall trend,the area of Haikou is divided into a grid of 1 km×1 km,and the empirical orthogonal function decomposition(EOF)method is introduced to decompose the travel time series into two parts:space modal and time coefficient.The results show that the overall travel volume showed an upward trend during the study period;And the travel patterns are highly cyclical;The spatial temporal patterns of travel are different between weekdays and weekends.Analyzing the travel patterns of residents can help to optimize the allocation of traffic resources.

关 键 词：网约车 上下车点数据 时间序列分解 经验正交函数 时空模式 

分 类 号：P208[天文地球—地图制图学与地理信息工程]