基于轨迹数据场的热点区域提取及空间交互分析——以深圳市为例
|
摘要
以深圳市出租车GPS数据为基础,运用时空拓展的轨迹数据场聚类方法提取城市交通热点区域,结合城市POI(Point of Interest)数据和地理实况对热点区域加以理解和分析。基于复杂网络的视角,计算交互分析指标并可视化热点区域的空间交互网络,探究城市交通和居民出行的时空规律。结果表明:1)交通枢纽(机场、火车站和口岸)、综合性商圈、城市重要主干道周边和城市商务中心在节假日和工作日均表现为持续热点区域;2)节假日热点区域分布较"发散",主要反映了居民个性化出行需求;3)工作日热点区域分布较"收敛",主要表现为职住分离的通勤模式;4)不同热点区域在空间交互网络中的重要性存在明显差异,其空间交互体现了距离衰减效应和局部抱团现象,居民出行的热点区域网络本身具有小世界效应和无标度特征。
As a kind of important public transportation, taxi trajectory data contain abundant information about urban functions and citizen activities owing to its long service time, wide coverage of city area and freedom of its motion. Based on taxi GPS data of Shenzhen city, in this study, we extract urban traffic hotspots using the clustering method of trajectory data field based on space-time development. The hotspot area spatio-temporal distribution of typical time periods of the two time types of holidays and workdays are selected for visual graphic representation, and analyze hot spots combining POI data and the present situation of urban development. Based on the view of complex network, the interaction analysis index is calculated and the spatial interaction network of hot spots is visualized, a qualitative and quantitative analysis method is used to explore the spatial and temporal rules of urban traffic and residents' travel. The results are showed as follows. 1) The hotspot area spatio-temporal distribution of urban traffic hotspots in holidays and workdays is significantly different. Transportation hub(like airports, railway stations and ports), a comprehensive business circle, area around urban major road and central business district are continuous hot spots during holidays and workdays, and the locations of other hotspots change over time. 2) The distribution of hot spots in holidays is disperse, mainly reflects the personalized travel demands of residents. Urban tourist attractions and other leisure areas have become hot spots, and it is clear that that outflow of Baoan airport and Shenzhen north station are obviously great than the inflow in the holiday season, which means that they are relatively strong. 3) The distribution of hot spots in workdays is convergent, which reflects the commuting mode of separation of work and residence. Residents' travel shows a tidal movement between the work place and the residence, the synchronism of inflow and outflow in hot spots is stronger than that in holidays, and residents travel more regularly. 4) Different hotspots in space interaction network have obviously difference in their significance, which spatial interaction reflects the distance-decay effect and partial aggregation phenomenon. Shennan avenue has become the dividing line between hot spots with close connections. Moreover, it was found that that space interaction network between Shenzhen north station, Futian district and Luohu district is of high importance in the space interaction network, and the links between the three are frequent. The hotspots network of residents travel have the characteristics of small-world effect and scale-free.
As a kind of important public transportation, taxi trajectory data contain abundant information about urban functions and citizen activities owing to its long service time, wide coverage of city area and freedom of its motion. Based on taxi GPS data of Shenzhen city, in this study, we extract urban traffic hotspots using the clustering method of trajectory data field based on space-time development. The hotspot area spatio-temporal distribution of typical time periods of the two time types of holidays and workdays are selected for visual graphic representation, and analyze hot spots combining POI data and the present situation of urban development. Based on the view of complex network, the interaction analysis index is calculated and the spatial interaction network of hot spots is visualized, a qualitative and quantitative analysis method is used to explore the spatial and temporal rules of urban traffic and residents' travel. The results are showed as follows. 1) The hotspot area spatio-temporal distribution of urban traffic hotspots in holidays and workdays is significantly different. Transportation hub(like airports, railway stations and ports), a comprehensive business circle, area around urban major road and central business district are continuous hot spots during holidays and workdays, and the locations of other hotspots change over time. 2) The distribution of hot spots in holidays is disperse, mainly reflects the personalized travel demands of residents. Urban tourist attractions and other leisure areas have become hot spots, and it is clear that that outflow of Baoan airport and Shenzhen north station are obviously great than the inflow in the holiday season, which means that they are relatively strong. 3) The distribution of hot spots in workdays is convergent, which reflects the commuting mode of separation of work and residence. Residents' travel shows a tidal movement between the work place and the residence, the synchronism of inflow and outflow in hot spots is stronger than that in holidays, and residents travel more regularly. 4) Different hotspots in space interaction network have obviously difference in their significance, which spatial interaction reflects the distance-decay effect and partial aggregation phenomenon. Shennan avenue has become the dividing line between hot spots with close connections. Moreover, it was found that that space interaction network between Shenzhen north station, Futian district and Luohu district is of high importance in the space interaction network, and the links between the three are frequent. The hotspots network of residents travel have the characteristics of small-world effect and scale-free.
引文
Barabási A L and Albert R.1999.Emergence of scaling in random networks.Science,286:509-512.
Chang Hanwen,Tai Yuchin and Hsu J Y J.2010.Context-aware taxi demand hotspots prediction.International Journal of Business Intelligence and Data Mining,5(1):3-18.
陈世莉,陶海燕,李旭亮,卓莉.2016.基于潜在语义信息的城市功能区识别--广州市浮动车GPS时空数据挖掘.地理学报,71(3):471-483.[Chen Shili,Tao Haiyan,Li Xuliang and Zhuo Li.2016.Discovering urban functional regions using latent semantic information:Spatiotemporal data mining of floating cars GPS data of Guangzhou.Acta Geographica Sinica,71(3):471-483.]
淦文燕,李德毅,王建民.2006.一种基于数据场的层次聚类方法.电子学报,(2):258-262.[Gan Wenyan,Li Deyi and Wang Jianmin.2006.An Hierarchical Clustering Method Based on Data Fields.Acta Electronica Sinica,(2):258-262.]
龚玺,裴韬,孙嘉,罗明.2011.时空轨迹聚类方法研究进展.地理科学进展,30(5):522-534.[Gong Xi,Pei Tao,Sun Jia and Luo Ming.2011.Review of the Research Progresses in Trajectory Clustering Methods.Progress in Geography,30(5):522-534.]
桂智明,向宇,李玉鉴.2012.基于出租车轨迹的并行城市热点区域发现.华中科技大学学报(自然科学版),40(S1):187-190.[Gui Zhiming,Xiang Yu and Li Yujian.2012.Parallel discovering of city hot spot based on taxi trajectories.Journal of Huazhong University of Science and Technology(Natural Science Edition),40(S1):187-190.]
Jahnke M,Ding L,Karja K and Wang Shirui.2017.Identifying Origin/Destination Hotspots in Floating Car Data for Visual Analysis of Traveling Behavior.In:GartnerGeorg and Huang Haosheng.Progress in Location-Based Services 2016.Berlin:Springer International Publishing,253-269.
李德毅,杜鹢.2005.不确定性人工智能.北京:国防工业出版社.[Li Deyi and Du Yi.2005.Artificial Intelligence with uncertainty.Beijing:National Defense Industry Press.]
李德毅,淦文燕,刘璐莹.2003.人工智能与认知物理学//中国人工智能学会第10届全国学术年会论文集.北京:北京邮电大学出版社,6-15.[Li Deyi,Gan Wenyan and Liu Luying.Artificial intelligence and cognitive physics//Progress of Artificial Intelligence in China.Beijing:BUPT Publishing House,6-15.]
李婷,裴韬,袁烨城,宋辞,王维一,杨格格.2014.人类活动轨迹的分类、模式和应用研究综述.地理科学进展,33(7):938-948.[Li Ting,Pei Tao,Yuan Yecheng,Song Ci,Wang Weiyi and Yang Gege.2014.A review on the classification,patterns and applied research of human mobility trajectory.Progress in Geography,33(7):938-948.]
李晓丹,杨晓光,陈华杰.2009.城市道路网络交通小区划分方法研究.计算机工程与应用,45(5):19-22.[Li Xiaodan,Yang Xiaoguang and Chen Huajie.2009.Study on traffic zone division based on spatial clustering analysis.Computer Engineering and Applications,45(5):19-22.]
牟乃夏,张恒才,陈洁,张灵先,戴洪磊.2015.轨迹数据挖掘城市应用研究综述.地球信息科学学报,17(10):1136-1142.[Mou Naixia,Zhang Hengcai,Chen Jie,Zhang Lingxian and Dai Honglei.2015.AReview on the Application Research of Trajectory Data Mining in Urban Cities.Journal of Geo-Information Science,17(10):1136-1142.]
Miller H J.2004.Tobler’s First Law and spatial analysis.Annals of the Association of American Geographers,94(2):284-289.
秦昆,周勍,徐源泉,徐雯婷,罗萍.2017.城市交通热点区域的空间交互网络分析.地理科学进展,36(9):1149-1157.[Qin Kun,Zhou Qing,Xu Yuanquan,Xu Wenting and Luo Ping.2017.Spatial interaction network analysis of urban traffic hotspots.Progress in Geography,36(9):1149-1157.]
Tobler W.1970.A computer movie simulating urban growth in the Detroit region.Economic Geography,46(2):234-240.
Ullman E.1941.A theory of location for cities.American Journal of Sociology,46(6):853-864.
Watts D J and Strogatz S H.1998.Collective dynamics of‘smallword’network.Nature,393:440-442.
Zhang Pengdong,Deng Min,Shi Yan and Zhao Ling.2016.Detecting hotspots of urban residents’behaviors based on spatio-temporal clustering techniques.GeoJournal,82(5):1-13.
Zhao Pengxiang,Qin Kun,Ye Xinyue and Chen Yixiang.2017.A trajectory clustering approach based on decision graph and data field for detecting hotspots.International Journal of Geographical Information Science,31(6):1101-1127.
Zheng Yu,Zhang Lizhu,Xie Xing and Ma Weiying.2009.Mining interesting locations and travel sequences from GPS trajectories.Proceedings of the 18th international conference on world wide web.Madrid,Spain:ACM:791-800.
赵鹏祥.2015.基于轨迹聚类的城市热点区域提取与分析方法研究.武汉:武汉大学.[Zhao Pengxiang.2015.Research on the Method of Extracting and Analyzing Urban Hotspots Based on Trajectory Clustering.Wuhan:Wuha University.]
周勍,秦昆,陈一祥,李志鑫.2016.基于数据场的出租车轨迹热点区域探测方法.地理与地理信息科学,32(6):51-56,127.[Zhou Qing,Qin Kun,Chen Yixiang and Li Zhixin.2016.Hotspots Detection from Taxi Trajectory Data Based on Data Field Clustering.Geography and Geo-Information Science,32(6):51-56,127.]
(1)百度地图POI行业分类:http://lbsyun.baidu.com/index.php?title=lbscloud/poitags。
Chang Hanwen,Tai Yuchin and Hsu J Y J.2010.Context-aware taxi demand hotspots prediction.International Journal of Business Intelligence and Data Mining,5(1):3-18.
陈世莉,陶海燕,李旭亮,卓莉.2016.基于潜在语义信息的城市功能区识别--广州市浮动车GPS时空数据挖掘.地理学报,71(3):471-483.[Chen Shili,Tao Haiyan,Li Xuliang and Zhuo Li.2016.Discovering urban functional regions using latent semantic information:Spatiotemporal data mining of floating cars GPS data of Guangzhou.Acta Geographica Sinica,71(3):471-483.]
淦文燕,李德毅,王建民.2006.一种基于数据场的层次聚类方法.电子学报,(2):258-262.[Gan Wenyan,Li Deyi and Wang Jianmin.2006.An Hierarchical Clustering Method Based on Data Fields.Acta Electronica Sinica,(2):258-262.]
龚玺,裴韬,孙嘉,罗明.2011.时空轨迹聚类方法研究进展.地理科学进展,30(5):522-534.[Gong Xi,Pei Tao,Sun Jia and Luo Ming.2011.Review of the Research Progresses in Trajectory Clustering Methods.Progress in Geography,30(5):522-534.]
桂智明,向宇,李玉鉴.2012.基于出租车轨迹的并行城市热点区域发现.华中科技大学学报(自然科学版),40(S1):187-190.[Gui Zhiming,Xiang Yu and Li Yujian.2012.Parallel discovering of city hot spot based on taxi trajectories.Journal of Huazhong University of Science and Technology(Natural Science Edition),40(S1):187-190.]
Jahnke M,Ding L,Karja K and Wang Shirui.2017.Identifying Origin/Destination Hotspots in Floating Car Data for Visual Analysis of Traveling Behavior.In:GartnerGeorg and Huang Haosheng.Progress in Location-Based Services 2016.Berlin:Springer International Publishing,253-269.
李德毅,杜鹢.2005.不确定性人工智能.北京:国防工业出版社.[Li Deyi and Du Yi.2005.Artificial Intelligence with uncertainty.Beijing:National Defense Industry Press.]
李德毅,淦文燕,刘璐莹.2003.人工智能与认知物理学//中国人工智能学会第10届全国学术年会论文集.北京:北京邮电大学出版社,6-15.[Li Deyi,Gan Wenyan and Liu Luying.Artificial intelligence and cognitive physics//Progress of Artificial Intelligence in China.Beijing:BUPT Publishing House,6-15.]
李婷,裴韬,袁烨城,宋辞,王维一,杨格格.2014.人类活动轨迹的分类、模式和应用研究综述.地理科学进展,33(7):938-948.[Li Ting,Pei Tao,Yuan Yecheng,Song Ci,Wang Weiyi and Yang Gege.2014.A review on the classification,patterns and applied research of human mobility trajectory.Progress in Geography,33(7):938-948.]
李晓丹,杨晓光,陈华杰.2009.城市道路网络交通小区划分方法研究.计算机工程与应用,45(5):19-22.[Li Xiaodan,Yang Xiaoguang and Chen Huajie.2009.Study on traffic zone division based on spatial clustering analysis.Computer Engineering and Applications,45(5):19-22.]
牟乃夏,张恒才,陈洁,张灵先,戴洪磊.2015.轨迹数据挖掘城市应用研究综述.地球信息科学学报,17(10):1136-1142.[Mou Naixia,Zhang Hengcai,Chen Jie,Zhang Lingxian and Dai Honglei.2015.AReview on the Application Research of Trajectory Data Mining in Urban Cities.Journal of Geo-Information Science,17(10):1136-1142.]
Miller H J.2004.Tobler’s First Law and spatial analysis.Annals of the Association of American Geographers,94(2):284-289.
秦昆,周勍,徐源泉,徐雯婷,罗萍.2017.城市交通热点区域的空间交互网络分析.地理科学进展,36(9):1149-1157.[Qin Kun,Zhou Qing,Xu Yuanquan,Xu Wenting and Luo Ping.2017.Spatial interaction network analysis of urban traffic hotspots.Progress in Geography,36(9):1149-1157.]
Tobler W.1970.A computer movie simulating urban growth in the Detroit region.Economic Geography,46(2):234-240.
Ullman E.1941.A theory of location for cities.American Journal of Sociology,46(6):853-864.
Watts D J and Strogatz S H.1998.Collective dynamics of‘smallword’network.Nature,393:440-442.
Zhang Pengdong,Deng Min,Shi Yan and Zhao Ling.2016.Detecting hotspots of urban residents’behaviors based on spatio-temporal clustering techniques.GeoJournal,82(5):1-13.
Zhao Pengxiang,Qin Kun,Ye Xinyue and Chen Yixiang.2017.A trajectory clustering approach based on decision graph and data field for detecting hotspots.International Journal of Geographical Information Science,31(6):1101-1127.
Zheng Yu,Zhang Lizhu,Xie Xing and Ma Weiying.2009.Mining interesting locations and travel sequences from GPS trajectories.Proceedings of the 18th international conference on world wide web.Madrid,Spain:ACM:791-800.
赵鹏祥.2015.基于轨迹聚类的城市热点区域提取与分析方法研究.武汉:武汉大学.[Zhao Pengxiang.2015.Research on the Method of Extracting and Analyzing Urban Hotspots Based on Trajectory Clustering.Wuhan:Wuha University.]
周勍,秦昆,陈一祥,李志鑫.2016.基于数据场的出租车轨迹热点区域探测方法.地理与地理信息科学,32(6):51-56,127.[Zhou Qing,Qin Kun,Chen Yixiang and Li Zhixin.2016.Hotspots Detection from Taxi Trajectory Data Based on Data Field Clustering.Geography and Geo-Information Science,32(6):51-56,127.]
(1)百度地图POI行业分类:http://lbsyun.baidu.com/index.php?title=lbscloud/poitags。