基于贝叶斯时空模型的城管事件时空变化分析

doi:10.12082/dqxxkx.2020.190413

地球信息科学学报 ›› 2020, Vol. 22 ›› Issue (5): 1073-1082.doi: 10.12082/dqxxkx.2020.190413

• “空间综合人文学与社会科学”专辑 • 上一篇下一篇

基于贝叶斯时空模型的城管事件时空变化分析

董文钱¹, 董良^2,³, 向琳^1,^*(), 陶海军¹, 赵传虎⁴, 曲寒冰^2,³

1.中国计量大学信息工程学院,杭州 310018
2.北京市科学技术研究院,北京 100089
3.北京市新技术应用研究所,北京 100094
4.河北工业大学人工智能与数据科学学院,天津 300401

收稿日期:2019-07-30 修回日期:2019-11-20 出版日期:2020-05-25 发布日期:2020-07-25
通讯作者: 向琳 E-mail:xianglin@cjlu.edu.cn
作者简介:董文钱（1995— ）,男,浙江温州人,硕士生,主要从事时空数据挖掘研究。E-mail: wqdong.chn@live.com
基金资助:
国家自然科学基金重大研究计划重点支持项目(NSF91746207);科技部国家重点研发计划项目(2018YFF0301000);科技部国家重点研发计划项目(2018YFC0809700);科技部国家重点研发计划项目(2018YFC0704800)

Spatiotemporal Variability of Urban Management Events based on the Bayesian Spatiotemporal Model

DONG Wenqian¹, DONG Liang^2,³, XIANG Lin^1,^*(), TAO Haijun¹, ZHAO Chuanhu⁴, QU Hanbing^2,³

1. College of Information Engineering, China Jiliang University, Hangzhou 310018, China
2. Beijing Academy of Science and Technology, Beijing 100089, China
3. Beijing Institute of New Technology Applications, Beijing 100094, China
4. School of Artificial Intelligence, Hebei University of Technology, Tianjin 300401, China

Received:2019-07-30 Revised:2019-11-20 Online:2020-05-25 Published:2020-07-25
Contact: XIANG Lin E-mail:xianglin@cjlu.edu.cn
Supported by:
National Natural Science Foundation of China(NSF91746207);National Key Research and Development Program of China(2018YFF0301000);National Key Research and Development Program of China(2018YFC0809700);National Key Research and Development Program of China(2018YFC0704800)

摘要/Abstract

摘要：

城市管理事件具有显著的地域特征和周期性特征,挖掘城管事件中隐含的时空规律和潜在影响因子对城市管理具有重要指导意义。然而,目前针对城管事件时空变化及影响因素驱动效应的研究仍较为少见。本文基于贝叶斯时空模型,针对西北某地H市P区的街面秩序类、市容环境类以及宣传广告类城管事件在一周之中的时空演变特征进行建模分析,并探究了影响城管事件案发风险的潜在影响因子。研究发现：① 在空间上,3类城管事件的相对风险分布存在差异,街面秩序类事件集中在城市的居住功能区和商业功能区,而市容环境类集中在城市的居住功能区,宣传广告类主要集中在城市的商业功能区,空间风险后验概率估计表明,以上2个区域是城管事件的热点区域。② 在时间上,每逢周二、周五以及周六,城管事件的相对风险较为突出,但总体上没有明显的单调性。每天的8—10时和14—15时是城管事件高发的时段,其相对风险远高于其他时段。③ 城市建成环境对城管事件的潜在影响存在差异。研究区域内餐饮、交通、生活服务等城市基础设施与城管事件的关联最为紧密,且都表现为正相关。④ 城管事件的案发风险呈现出明显的时空异质性,渐进性的建模过程表明在分析城管事件数据时考虑空间与时间效应的影响是合理且有必要的。研究表明,贝叶斯时空模型的分析结果满足城管部门日常工作中关于城管事件智能分析与重点管控的需求。

关键词: 城管事件, 贝叶斯时空模型, 结构效应, 非结构效应, 相对风险, 时空变化, 冷热点, INLA

Abstract:

Urban management events are regional and periodic. The spatiotemporal laws and potential impact factors implied in urban management events are vital for improving urban management. However, research on the temporal and spatial changes of urban management events and influencing factors are rare. In this paper, by using the Bayesian space-time model, we modeled and analyzed the temporal and spatial evolution characteristics of three types of city management events-street order, urban environment, and publicity advertising-in the P district of H city, Northwest China, and explored the impact of urban management events as well as the underlying impact factors. We found that: (1) There were spatial differences in the relative risk distribution of the three types of urban management events. The street order type was concentrated in the residential and commercial areas of the city, while the urban environment type was concentrated in the residential areas of the city. The advertising type was mainly concentrated in the commercial areas of the city. The spatial risk posterior probability estimate indicated that the above two regions are hotspots of urban management events. (2) The relative risks of urban management events were more prominent on Tuesdays, Fridays, and Saturdays, but there was no obvious monotony in general trends. Meanwhile, the hourly trends had irregular fluctuation, everyday from 8 to 10 and from 14 to 15, it was a period of the high incidence of urban management events, and its relative risk was much higher than other periods. (3) For different built environments, the potential impacts of these factors were quite different. The relative risk of urban management events was significantly associated with restaurants, transportation, and living services, all positively correlated. (4) The relative risk of urban management events presented obvious spatial and temporal heterogeneity. and it is reasonable and necessary to consider the impact of spatial and temporal effects when analyzing urban management events data. Our findings are meaningful for relavant government departments to make effective policies to control and reduce the relative risk of urban management events, especially for the study area.

Key words: urban management events, Bayesian spatiotemporal model, structured effect, unstructured effect, relative risk, spatiotemporal variability, hot/cold spots, INLA

董文钱, 董良, 向琳, 陶海军, 赵传虎, 曲寒冰. 基于贝叶斯时空模型的城管事件时空变化分析[J]. 地球信息科学学报, 2020, 22(5): 1073-1082.DOI:10.12082/dqxxkx.2020.190413

DONG Wenqian, DONG Liang, XIANG Lin, TAO Haijun, ZHAO Chuanhu, QU Hanbing. Spatiotemporal Variability of Urban Management Events based on the Bayesian Spatiotemporal Model[J]. Journal of Geo-information Science, 2020, 22(5): 1073-1082.DOI:10.12082/dqxxkx.2020.190413

图/表 12

图1

图2

图3

表1

表2

不同组件的候选模型"

模型	模型结构
M1	$log λ ijk = b 0 + ∑ p β p X pijk + log E ij$
M2	$log λ ijk = b 0 + ∑ p β p X pijk + μ i + ν i + log E ij$
M3	$log λ ijk = b 0 + ∑ p β p X pijk + μ i + ν i + γ j + φ j + τ k + log E ij$

表2

表3

表4

图4

图5

图6

图7

表5

参考文献 31

[1]	王家耀. 大数据时代的智慧城市[J]. 测绘科学, 2014,39(5):3-7.
	[ Wang J Y. Smart city in the era of big data[J]. Science of Surveying and Mapping, 2014,39(5):3-7. ]
[2]	佘冰, 朱欣焰, 呙维, 等. 基于空间点模式分析的城市管理事件空间分布及演化:以武汉市江汉区为例[J]. 地理科学进展, 2013,32(6):924-931.
	[ She B, Zhu X Y, Guo W, et al. Spatial distribution and evolution of city management events based on the spatial point pattern analysis: A case study of Jianghan district, Wuhan city[J]. Progress in Geography, 2013,32(6):924-931. ]
[3]	靖常峰, 付艳丽, 戴培培, 等. 近邻指数的城管案件全局分布模式分析[J]. 测绘科学, 2017,42(6):76-81.
	[ Jing C F, Fu Y L, Dai P P, et al. Spatial pattern analysis on global distribution of urban management case based on nearest index[J]. Science of Surveying and Mapping, 2017,42(6):76-81. ]
[4]	伏家云, 靖常峰, 杜明义, 等. 参数优化DBSCAN算法的城管案件聚类分析[J]. 测绘科学, 2018,43(8):135-140.
	[ Fu J Y, Jing C F, Du M Y, et al. Urban management case spatial clustering analysis based on a parameter optimized DBSCAN algorithm[J]. Science of Surveying and Mapping, 2018,43(8):135-140. ]
[5]	Gelman A, Carlin J B, Stern H S, et al. Bayesian data analysis[M]. Boca Raton, FL: CRC Press, 2013.
[6]	Ugarte M D, Adin A, Goicoa T, et al. On fitting spatio-temporal disease mapping models using approximate Bayesian inference[J]. Statistical Methods in Medical Research, 2014,23(6):507-530.
[7]	Song C, Yang X, Shi X, et al. Estimating missing values in China’s official socioeconomic statistics using progressive spatiotemporal Bayesian hierarchical modeling[J]. Scientific Reports, 2018,8(1):10055. doi: 10.1038/s41598-018-28322-z
[8]	Lawson A B. Bayesian disease mapping: Hierarchical modeling in spatial epidemiology[M]. Boca Raton, FL: CRC Press, 2013.
[9]	Banerjee S, Carlin B P, Gelfand A E. Hierarchical modeling and analysis for spatial data[M]. Boca Raton, FL: CRC Press, 2014.
[10]	吴北平, 杨典, 王劲峰, 等. 利用贝叶斯时空模型分析山东省手足口病时空变化及影响因素[J]. 地球信息科学学报, 2016,18(12):1645-1652. doi: 10.3724/SP.J.1047.2016.01645
	[ Wu B P, Yang D, Wang J F, et al. Space-time variability and determinants of hand, foot and mouth in Shandong province: A Bayesian spatio-temporal modeling approach[J]. Journal of Geo-information Science, 2016,18(12):1645-1652. ] doi: 10.3724/SP.J.1047.2016.01645
[11]	Halonen J I, Blangiardo M, Toledano M B, et al. Is long-term exposure to traffic pollution associated with mortality? A small-area study in London[J]. Environmental Pollution, 2016,208:25-32. doi: 10.1016/j.envpol.2015.06.036
[12]	Osei F B, Stein A. Diarrhea morbidities in small areas: Accounting for non-stationarity in sociodemographic impacts using Bayesian spatially varying coefficient modelling[J]. Scientific Reports, 2017,7(1):9908. doi: 10.1038/s41598-017-10017-6
[13]	关雪峰, 曾宇媚. 时空大数据背景下并行数据处理分析挖掘的进展及趋势[J]. 地理科学进展, 2018,37(10):1314-1327. doi: 10.18306/dlkxjz.2018.10.002
	[ Guang X F, Zeng Y M. Research progress and trends of parallel processing, analysis, and mining of big spatiotemporal data[J]. Progress in Geography, 2018,37(10):1314-1327. ] doi: 10.18306/dlkxjz.2018.10.002
[14]	刘玲, 李钢, 杨兰, 等. 深圳市快递自提点的空间分布特征与影响因素[J]. 地球信息科学, 2019,21(8):1240-1253.
	[ Liu L, Li G, Yang L, et al. Spatial distribution characteristics and influencing factors of the delivery sites in Shenzhen[J]. Journal of Geo-information Science, 2019,21(8):1240-1253. ]
[15]	Vatcheva K P, Lee M J, McCormick J B, et al. Multicollinearity in regression analyses conducted in epidemiologic studies[J]. Epidemiology (Sunnyvale, Calif), 2016,6(2):227.
[16]	Besag J, York J, Mollie A. Bayesian image restoration, with two applications in spatial statistics[J]. Annals of the Institute of Statistical Mathematics, 1991,43(1):1-20.
[17]	Blangiardo M, Cameletti M. Spatial and spatio-temporal Bayesian models with R-INLA[M]. London: John Wiley and Sons, 2015.
[18]	Besag J. Spatial interaction and the statistical analysis of lattice systems[J]. Journal of the Royal Statistical Society: Series B(Methodological), 1974,36(2):192-225.
[19]	Lichstein J W, Simons T R, Shriner S A, et al. Spatial autocorrelation and autoregressive models in ecology[J]. Ecological Monographs, 2002,72(3):445-463.
[20]	Song C, He Y, Bo Y, et al. Risk assessment and mapping of hand, foot, and mouth disease at the county level in mainland China using spatiotemporal zero-inflated Bayesian hierarchical models[J]. International Journal of Environmental Research and Public Health, 2018,15(7):1476.
[21]	Wang Y, Lai Y, Du Z, et al. Spatiotemporal distribution of hand, foot, and mouth disease in Guangdong province, China and potential predictors, 2009-2012[J]. International Journal of Environmental Research and Public Health, 2019,16(7):1191.
[22]	王劲峰, 廖一兰, 刘鑫. 空间数据分析教程(第二版)[M]. 北京: 科学出版社, 2019.
	[ Wang J F, Liao Y L, Liu X. 2019. Spatial data analysis tutorial(second edition)[M]. Beijing: Science Press, 2019. ]
[23]	Schrodle B, Held L. Spatio-temporal disease mapping using INLA[J]. Environmetrics, 2011,22(6):725-734.
[24]	Spiegelhalter D J, Best N G, Carlin B P, et al. Bayesian measures of model complexity and fit[J]. Journal of the Royal Statistical Society:Series B(Statistical Methodology), 2002,64(4):583-639.
[25]	Rue H, Martino S, Chopin N. Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations[J]. Journal of the Royal Statistical Society: Series B(Statistical Methodology), 2009,71(2):319-392.
[26]	Richardson S, Thomson A, Best N, et al. Interpreting posterior relative risk estimates in disease-mapping studies[J]. Environmental Health Perspectives, 2004,112(9):1016-1025.
[27]	林文棋, 陈会宴, 谢盼, 等. 基于多源数据的北京市朝阳区人口时空格局评估与预测[J]. 地球信息科学学报, 2018,20(10):1467-1477.
	[ Lin W Q, Chen H Y, Xie P, et al. Spatial-temporal variation evaluation and prediction of population in Chaoyang district of Beijing based on multisource data[J]. Journal of Geo-information Science, 2018,20(10):1467-1477. ]
[28]	Knorr H L. Bayesian modelling of inseparable space-time variation in disease risk[J]. Statistics in Medicine, 2000,19(17-18):2555-2567.
[29]	Openshaw S. The modifiable areal unit problem//Concepts and Techniques in Modern Geography[M]. Norwich: Geo Books, 1984.
[30]	彭建, 徐飞雄. 不同格网尺度下的黄山市生境质量差异分析[J]. 地球信息科学学报, 2019,21(6):887-897.
	[ Peng J, Xu F X. Effect of grid size on habitat quality assessment: A case study of Huangshan city[J]. Journal of Geo-information Science, 2019,21(6):887-897. ]
[31]	李志林, 王继成, 谭诗腾, 等. 地理信息科学中尺度问题的30年研究现状[J]. 武汉大学学报·信息科学版, 2018,43(12):2233-2242.
	[ Li Z L, Wang J C, Tan S T, et al. Scale in Geo-information Science: An overview of thirty-year development[J]. Geomatics and Information Science of Wuhan University, 2018,43(12):2233-2242. ]

POI类别	POI内容	均值(标准差)
餐饮服务类	中餐厅、快餐厅、外国餐厅、休闲餐饮场所等	0.77(1.20)
交通设施类	公交车站、物流站等	0.20(0.37)
居民住宅类	住宅区	0.35(0.58)
购物服务类	便利店、超市、烟酒专卖店、购物相关场所等	0.71(1.07)
生活服务类	美容美发店、洗浴推拿场所、生活服务场所等	0.71(1.00)
住宿服务类	酒店、旅馆、招待所等	0.13(0.38)
医疗保健类	医院、社区诊所、卫生院、药店、药房等	0.37(0.69)
公司企业类	公司、企业、工厂等	0.68(0.78)

协变量(POI密度)	VIF	协变量(POI密度)	VIF
餐饮服务类	4.53	生活服务类	5.17
交通设施类	1.37	住宿服务类	2.52
居民住宅类	2.26	医疗保健类	3.21
购物服务类	3.88	公司企业类	1.41

协变量	街面秩序类	市容环境类	宣传广告类
截距项	0.586(0.549, 0.624)	0.654 (0.614, 0.697)	0.611 (0.573, 0.651)
餐饮服务类	1.083^*(1.032, 1.136)	1.065^*(1.017, 1.116)	1.086^* (1.036, 1.138)
交通设施类	1.105^*(1.047, 1.166)	1.056^* (1.001, 1.113)	1.091^* (1.035, 1.151)
居民住宅类	1.156^*(1.017, 1.315)	1.073 (0.949, 1.213)	1.118 (0.990, 1.263)
购物服务类	1.055(0.963, 1.157)	1.039 (0.950, 1.136)	1.052 (0.962, 1.150)
生活服务类	1.084^*(1.026, 1.145)	1.065^* (1.009, 1.125)	1.081^* (1.024, 1.141)
住宿服务类	1.158^*(1.019, 1.315)	1.201^* (1.070, 1.363)	1.205^* (1.069, 1.359)
医疗保健类	1.070(0.956, 1.162)	1.001 (0.931, 1.093)	1.081 (0.997, 1.171)
公司企业类	0.984(0.928, 1.043)	0.995 (0.941, 1.052)	0.944 (0.921, 1.031)

基于贝叶斯时空模型的城管事件时空变化分析

Spatiotemporal Variability of Urban Management Events based on the Bayesian Spatiotemporal Model

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 31

相关文章 15

Metrics

本文评价

推荐阅读 0

事件类别	候选模型	DIC
街面秩序	M1	36 873.15
	M2	26 182.07
	M3	24 851.99
市容环境	M1	27 093.02
	M2	21 791.38
	M3	19 099.85
宣传广告	M1	29 935.99
	M2	23 364.75
	M3	20 375.71

[1]	葛咏, 刘梦晓, 胡姗, 任周鹏. 时空统计学在贫困研究中的应用及展望[J]. 地球信息科学学报, 2021, 23(1): 58-74.
[2]	赵青松, 兰安军, 范泽孟, 杨青. 贵州省不同地貌形态类型土壤侵蚀强度变化的定量分析[J]. 地球信息科学学报, 2020, 22(7): 1555-1566.
[3]	叶虎平, 廖小罕, 何贤强, 岳焕印. 斯里兰卡近海海洋生态环境变化遥感监测分析[J]. 地球信息科学学报, 2020, 22(7): 1463-1475.
[4]	高旺旺, 冯建中, 白林燕, 杨建华, 郭雷风, 李华林, 崔梦瑞. 海南岛气溶胶时空变化及来源追溯[J]. 地球信息科学学报, 2020, 22(7): 1532-1543.
[5]	贾梦姝, 张宇, 潘婷婷, 吴文周, 苏奋振. 面向互联网信息抽取的海洋环境灾害链本体构建——以台风灾害为例[J]. 地球信息科学学报, 2020, 22(12): 2289-2303.
[6]	周玲, 李云梅, 赵焕, 房晟忠, 颜翔, 吕恒. 滇池湖滨带植被时空变化及其对滇池总磷浓度的影响[J]. 地球信息科学学报, 2020, 22(10): 1946-1958.
[7]	洪恬林, 李云梅, 吕恒, 孟斌, 毕顺, 周玲. 基于MODIS数据的太湖浮游植物物候变化及其对水表温度的响应[J]. 地球信息科学学报, 2020, 22(10): 1935-1945.
[8]	高黎明,张乐乐. 青海湖流域植被盖度时空变化研究[J]. 地球信息科学学报, 2019, 21(9): 1318-1329.
[9]	李健锋, 叶虎平, 张宗科, 孔金玲, 魏显虎, Somasundaram Deepakrishna, 王法溧. 基于Landsat影像的斯里兰卡内陆湖库水体时空变化分析[J]. 地球信息科学学报, 2019, 21(5): 781-788.
[10]	李娅丽, 汪小钦, 陈芸芝, 王苗苗. 福建省地表温度与植被覆盖度的相关性分析[J]. 地球信息科学学报, 2019, 21(3): 445-454.
[11]	郑海亮, 房世峰, 刘成程, 吴金华, 杜加强. 青藏高原月NDVI时空动态变化及其对气候变化的响应[J]. 地球信息科学学报, 2019, 21(2): 201-214.
[12]	徐轩, 李均力, 包安明, 王宝山, 李长春. 新疆五彩湾矿区开发对荒漠植被的扰动分析[J]. 地球信息科学学报, 2019, 21(12): 1934-1944.
[13]	牛丽楠, 邵全琴, 刘国波, 唐玉芝. 六盘水市土壤侵蚀时空特征及影响因素分析[J]. 地球信息科学学报, 2019, 21(11): 1755-1767.
[14]	傅俐,王勇,曾彪,毛泳,高敏. 基于改进两步移动搜索法的北碚区医疗设施空间可达性分析[J]. 地球信息科学学报, 2019, 21(10): 1565-1575.
[15]	洪娅岚,薛存金,刘敬一,刘星,孙强,伍程斌. 全球海洋初级生产力时空异常变化对ENSO事件的响应[J]. 地球信息科学学报, 2019, 21(10): 1538-1549.