地球信息科学学报 ›› 2021, Vol. 23 ›› Issue (11): 1924-1925.doi: 10.12082/dqxxkx.2021.210576
• 专栏:全球新型冠状病毒肺炎(COVID-19)疫情时空建模与决策分析 • 上一篇 下一篇
收稿日期:
2021-09-25
修回日期:
2021-11-09
出版日期:
2021-11-25
发布日期:
2022-01-25
通讯作者:
*高惠瑛(1967— ),女,山东青岛人,博士,教授,主要从事城市灾害的风险管理以及城市安全管理信息系统研发。 E-mail: fqmghy@sina.com作者简介:
李照(1996— ),女,山东济南人,硕士生,主要从事城市突发公共卫生防灾及风险评估研究。E-mail: zz2015@stu.ouc.edu.cn
基金资助:
LI Zhao(), GAO Huiying(
), DAI Xiaoyi, SUN Hai
Received:
2021-09-25
Revised:
2021-11-09
Online:
2021-11-25
Published:
2022-01-25
Contact:
*GAO Huiying, E-mail: fqmghy@sina.comSupported by:
摘要:
模拟传染病时空传播、定量评估疫情风险对科学防控、精准施策具有重要的现实意义。本文融合多源时空数据,构建了耦合LSTM算法和云模型的疫情传播风险预测模型。该模型首先基于GIS和LSTM算法构建疫情空间演变模拟模型,通过学习历史疫情数据中的规律,以1 km×1 km为空间尺度、天为时间尺度模拟传染病时空传播过程。其次,基于模拟传染病例数据和疫情传播时空影响因素构建风险评价指标,应用云模型和自适应策略构建疫情风险评估模型,实现多空间尺度的疫情风险评价。在实证研究阶段,应用该模型对北京2020年6月份突发COVID-19疫情空间演变过程进行模拟和风险评估,并引入常规机器学习模型作比较验证。结果表明:应用于疫情时空传播模拟,相较其它常规的机器学习模型,考虑时序关系的LSTM模型的模拟精度更高(MAE为0.00261),拟合度更好(R-square为0.9455);耦合模型不仅能充分考虑传染源因素、天气因素、疫情扩散因素及疫情防御因素对疫情风险传播的影响,反映风险演变趋势,还能快速量化区域风险等级,实现不同空间分辨率下的疫情风险评估。因此,基于LSTM算法和云模型的耦合模型可有效预测疫情的传播风险,同时,也为传染病时空传播建模与风险评估提供了方法参考。
李照, 高惠瑛, 代晓奕, 孙海. 一种耦合LSTM算法和云模型的疫情传播风险预测模型[J]. 地球信息科学学报, 2021, 23(11): 1924-1925.DOI:10.12082/dqxxkx.2021.210576
LI Zhao, GAO Huiying, DAI Xiaoyi, SUN Hai. An Epidemic Spread Risk Prediction Model Coupled with LSTM Algorithm and Cloud Model[J]. Journal of Geo-information Science, 2021, 23(11): 1924-1925.DOI:10.12082/dqxxkx.2021.210576
表2
实验数据
影响因素 | 基础数据 | 来源 | 数据类型 |
---|---|---|---|
疫情 | 新增确诊病例(2020-06-11—2020-07-01,共309条) | 北京卫健委 | 矢量(点) |
天气 | 2 m处最高温度数据、2 m处最低温度数据、总降水量数据、总天空直接太阳辐射量数据 | 欧洲气象中心发布的ERA5资料[ | 0.125°分辨率栅格 |
人口流动 | 基于微博签到数据的区人口流动指数[ | 新浪微博发布的位置信息 | 矢量(面) |
人口聚集 | 百度热力图(2020-06-11—2020-07-01) | 百度地图APP | 1 km×1 km栅格 |
政策 | 乡镇(街道)区域风险等级 | 北京卫健委、北京市疾病预防控制中心 | 矢量(面) |
表3
隐藏层调整的模拟结果
隐藏层神 经元个数 | MAE (6-25—6-28) | MAE (6-29—7-1) | MAE 合计 |
---|---|---|---|
3 | 0.002 06 | 0.001 65 | 0.003 71 |
4 | 0.002 14 | 0.001 54 | 0.003 68 |
5 | 0.002 12 | 0.001 69 | 0.003 81 |
6 | 0.002 05 | 0.001 31 | 0.003 36 |
7 | 0.001 86 | 0.001 23 | 0.003 09 |
8 | 0.001 85 | 0.001 02 | 0.002 87 |
9 | 0.001 83 | 0.001 13 | 0.002 96 |
10 | 0.001 91 | 0.001 27 | 0.003 18 |
11 | 0.002 12 | 0.001 22 | 0.003 34 |
12 | 0.002 05 | 0.001 17 | 0.003 22 |
表6
云模型的计算参数矩阵
指标 | 低风险 | 较低风险 | 中风险 | 较高风险 | 高风险 |
---|---|---|---|---|---|
A1 | (0.0005, 0.0003, 0.1) | (0.0030, 0.0013, 0.1) | (0.0065, 0.0010, 0.1) | (0.0540, 0.0307, 0.1) | (0.1100, 0.0067, 0.1) |
B1 | (146.42, 97.61, 0.1) | (293.97, 0.76, 0.1) | (296.08, 0.65, 0.1) | (298.14, 0.72, 0.1) | (300.15, 0.62, 0.1) |
B2 | (144.63, 96.42, 0.01) | (290.35, 0.72, 0.01) | (292.29, 0.57, 0.01) | (294.10, 0.64, 0.01) | (295.83, 0.52, 0.01) |
B3 | (0.0008, 0.0005, 0.1) | (0.0020, 0.0002, 0.1) | (0.0029, 0.0003, 0.1) | (0.0041, 0.0004, 0.1) | (0.005 9, 0.000 8, 0.1) |
B4 | (1 288 026, 858 684,0.1) | (2 691 607, 77037, 0.1) | (2 902 592, 63620, 0.1) | (3120 919, 81 931, 0.1) | (3 379 396, 90 387, 0.1) |
C1 | (0.1500, 0.1000, 0.01) | (0.4000, 0.0667, 0.01) | (0.6000, 0.0667, 0.01) | (0.7500, 0.0333, 0.01) | (0.850 0, 0.0333, 0.01) |
D1 | (0.5000, 0.3333, 0.01) | (1.5000, 0.3333, 0.01) | (2.5000, 0.3333, 0.01) | (4.0000, 0.6667, 0.01) | (6.000 0, 0.6667, 0.01) |
D2 | (0.4409, 0.2940, 0.01) | (1.5724, 0.4603, 0.01) | (3.1944, 0.6210, 0.01) | (5.3173, 0.7942, 0.01) | (8.2675, 1.1726, 0.01) |
[1] | Dogan O, Tiwari S, Jabbar M A, et al. A systematic review on AI/ML approaches against COVID-19 outbreak[J]. Complex & Intelligent Systems, 2021, 7(5):2655-2678. |
[2] |
裴韬, 王席, 宋辞, 等. COVID-19疫情时空分析与建模研究进展[J]. 地球信息科学学报, 2021, 23(2):188-210.
doi: 10.12082/dqxxkx.2021.200434 |
[ Pei T, Wang X, Song C, et al. Review on spatiotemporal analysis and modeling of COVID-19 pandemic[J]. Journal of Geo- information Science, 2021, 23(2):188-210. ] | |
[3] |
Hilker F M, Langlais M, Petrovskii S V, et al. A diffusive SI model with Allee effect and application to FIV[J]. Mathematical Biosciences, 2007, 206(1):61-80.
pmid: 16387332 |
[4] |
Van Mieghem P. The N-intertwined SIS epidemic network model[J]. Computing, 2011, 93(2/3/4):147-169.
doi: 10.1007/s00607-011-0155-y |
[5] |
Bjørnstad O N, Finkenstädt B F, Grenfell B T. Dynamics of measles epidemics: Estimating scaling of transmission rates using a time series sir model[J]. Ecological Monographs, 2002, 72(2):169-184.
doi: 10.1890/0012-9615(2002)072[0169:DOMEES]2.0.CO;2 |
[6] |
Das A, Dhar A, Goyal S, et al. COVID-19: Analytic results for a modified SEIR model and comparison of different intervention strategies[J]. Chaos Solitons & Fractals, 2021, 144(2):110595.
doi: 10.1016/j.chaos.2020.110595 |
[7] |
López L, Rodó X. A modified SEIR model to predict the COVID-19 outbreak in Spain and Italy: Simulating control scenarios and multi-scale epidemics[J]. Results in Physics, 2021, 21:103746.
doi: 10.1016/j.rinp.2020.103746 |
[8] |
毕佳, 王贤敏, 胡跃译, 等. 一种基于改进SEIR模型的突发公共卫生事件风险动态评估与预测方法—以欧洲十国COVID-19为例[J]. 地球信息科学学报, 2021, 23(2):259-273.
doi: 10.12082/dqxxkx.2021.200356 |
[ Bi J, Wang X M, Hu Y Y, et al. A method for dynamic risk assessment and prediction of public health emergencies based on an improved SEIR model: COVID-19 in ten European countries[J]. Journal of Geo-information Science, 2021, 23(2):259-273. ] | |
[9] | 夏吉喆, 周颖, 李珍, 等. 城市时空大数据驱动的新型冠状病毒传播风险评估——以粤港澳大湾区为例[J]. 测绘学报, 2020, 49(6):671-680. |
[ Xia J Z, Zhou Y, Li Z, et al. COVID-19 risk assessment driven by urban spatiotemporal big data: A case study of Guangdong-Hong Kong-Macao Greater Bay Area[J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(6):671-680. ] | |
[10] | 冯明翔, 方志祥, 路雄博, 等. 交通分析区尺度上的新型冠状病毒肺炎时空扩散推估方法:以武汉市为例[J]. 武汉大学学报·信息科学版, 2020, 45(5):651-657,681. |
[ Feng M X, Fang Z X, Lu X B, et al. Traffic analysis zone-based epidemic estimation approach of COVID-19 based on mobile phone data: an example of Wuhan[J/OL]. Geomatics and Information Science of Wuhan University, 2020, 45(5):651-657,681. ] | |
[11] |
Koo J R, Cook A R, Park M, et al. Interventions to mitigate early spread of SARS-CoV-2 in Singapore: A modelling study[J]. The Lancet Infectious Diseases, 2020, 20(6):678-688.
doi: 10.1016/S1473-3099(20)30162-6 |
[12] |
Brooks L C, Farrow D C, Hyun S, et al. Flexible modeling of epidemics with an empirical Bayes framework[J]. PLoS Computational Biology, 2015, 11(8):e1004382.
doi: 10.1371/journal.pcbi.1004382 |
[13] |
Benvenuto D, Giovanetti M, Vassallo L, et al. Application of the ARIMA model on the COVID-2019 epidemic dataset[J]. Data in Brief, 2020, 29:105340.
doi: 10.1016/j.dib.2020.105340 pmid: 32181302 |
[14] |
Kufel T. ARIMA-based forecasting of the dynamics of confirmed Covid-19 cases for selected European countries[J]. Equilibrium, 2020, 15(2):181-204.
doi: 10.24136/eq.2020.009 |
[15] |
Malavika B, Marimuthu S, Joy M, et al. Forecasting COVID-19 epidemic in India and high incidence states using SIR and logistic growth models[J]. Clinical Epidemiology and Global Health, 2021, 9:26-33.
doi: 10.1016/j.cegh.2020.06.006 |
[16] |
Jiang D, Hao M M, Ding F Y, et al. Mapping the transmission risk of Zika virus using machine learning models[J]. Acta Tropica, 2018, 185:391-399.
doi: S0001-706X(18)30361-9 pmid: 29932934 |
[17] |
Adamker G, Holzer T, Karakis I, et al. Prediction of Shigellosis outcomes in Israel using machine learning classifiers[J]. Epidemiology and Infection, 2018, 146(11):1445-1451.
doi: 10.1017/S0950268818001498 pmid: 29880081 |
[18] |
Chimmula V K R, Zhang L. Time series forecasting of COVID-19 transmission in Canada using LSTM networks[J]. Chaos, Solitons & Fractals, 2020, 135:109864.
doi: 10.1016/j.chaos.2020.109864 |
[19] | Gao Z W, Lang M. Design of H7N9 avian influenza management and forecasting system based on GIS [C]//2015 IEEE 5th International Conference on Electronics Information and Emergency Communication. IEEE, 2015:376-379. |
[20] | 任红艳, 吴伟, 李乔玄, 等. 基于反向传播神经网络模型的广东省登革热疫情预测研究[J]. 中国媒介生物学及控制杂志, 2018, 29(3):221-225. |
[ Ren H Y, Wu W, Li Q X, et al. Prediction of dengue fever based on back propagation neural network model in Guangdong, China[J]. Chinese Journal of Vector Biology and Control, 2018, 29(3):221-225. ] | |
[21] | 李卫红, 陈业滨, 闻磊. 基于GA-BP神经网络模型的登革热时空扩散模拟[J]. 中国图象图形学报, 2015, 20(7):981-991. |
[ Li W H, Chen Y B, Wen L. Simulation of spatio-temporal diffusion of dengue fever based on the GA-BP neural network model[J]. Journal of Image and Graphics, 2015, 20(7):981-991. ] | |
[22] | 陈业滨, 李卫红. 支持向量机模型的登革热时空扩散预测[J]. 测绘科学, 2017, 42(2):65-70. |
[ Chen Y B, Li W H. Simulation of spatio-temporal diffusion trend of dengue fever based on the SVM model[J]. Science of Surveying and Mapping, 2017, 42(2):65-70. ] | |
[23] |
Pourghasemi H R, Pouyan S, Farajzadeh Z, et al. Assessment of the outbreak risk, mapping and infection behavior of COVID-19: Application of the autoregressive integrated-moving average (ARIMA) and polynomial models[J]. PLoS One, 2020, 15(7):e0236238.
doi: 10.1371/journal.pone.0236238 |
[24] |
Ong J, Liu X, Rajarethinam J, et al. Mapping dengue risk in Singapore using Random Forest[J]. PLoS Neglected Tropical Diseases, 2018, 12(6):e0006587.
doi: 10.1371/journal.pntd.0006587 |
[25] |
Liang R R, Lu Y, Qu X S, et al. Prediction for global African swine fever outbreaks based on a combination of random forest algorithms and meteorological data[J]. Transboundary and Emerging Diseases, 2020, 67(2):935-946.
doi: 10.1111/tbed.v67.2 |
[26] |
宋关福, 陈勇, 罗强, 等. GIS基础软件技术体系发展及展望[J]. 地球信息科学学报, 2021, 23(1):2-15.
doi: 10.12082/dqxxkx.2021.210015 |
[ Song G F, Chen Y, Luo Q A, et al. Development and prospect of GIS platform software technology system[J]. Journal of Geo-information Science, 2021, 23(1):2-15. ] | |
[27] |
Xie Z X, Yan J. Kernel Density Estimation of traffic accidents in a network space[J]. Computers, Environment and Urban Systems, 2008, 32(5):396-406.
doi: 10.1016/j.compenvurbsys.2008.05.001 |
[28] | 王鑫, 吴际, 刘超, 等. 基于LSTM循环神经网络的故障时间序列预测[J]. 北京航空航天大学学报, 2018, 44(4):772-784. |
[ Wang X, Wu J, Liu C, et al. Exploring LSTM based recurrent neural network for failure time series prediction[J]. Journal of Beijing University of Aeronautics and Astronautics, 2018, 44(4):772-784. ] | |
[29] |
Hochreiter S, Schmidhuber J. Long short-term memory[J]. Neural Computation, 1997, 9(8):1735-1780.
pmid: 9377276 |
[30] |
Peng T, Deng H W. Comprehensive evaluation for sustainable development based on relative resource carrying capacity-a case study of Guiyang, Southwest China[J]. Environmental Science and Pollution Research International, 2020, 27(16):20090-20103.
doi: 10.1007/s11356-020-08426-8 pmid: 32236806 |
[31] |
Byass P. Eco-epidemiological assessment of the COVID-19 epidemic in China, January-February 2020[J]. Global Health Action, 2020, 13(1):1760490.
doi: 10.1080/16549716.2020.1760490 |
[32] |
Paez A, Lopez F A, Menezes T, et al. A spatio-temporal analysis of the environmental correlates of COVID-19 incidence in Spain[J]. Geographical Analysis, 2021, 53(3):397-421.
doi: 10.1111/gean.v53.3 |
[33] |
Jia J S, Lu X, Yuan Y, et al. Population flow drives spatio-temporal distribution of COVID-19 in China[J]. Nature, 2020, 582(7812):389-394.
doi: 10.1038/s41586-020-2284-y |
[34] |
Zhu Y J, Xie J G, Huang F M, et al. The mediating effect of air quality on the association between human mobility and COVID-19 infection in China[J]. Environmental Research, 2020, 189:109911.
doi: 10.1016/j.envres.2020.109911 |
[35] | Copernicus Climate Change Service (C3S)[DB/OL]. https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-single-levels?tab=form. |
[36] |
潘碧麟, 王江浩, 葛咏, 等. 基于微博签到数据的成渝城市群空间结构及其城际人口流动研究[J]. 地球信息科学学报, 2019, 21(1):68-76.
doi: 10.12082/dqxxkx.2019.180235 |
[ Pan B L, Wang J H, Ge Y, et al. Spatial structure and population flow analysis in Chengdu-Chongqing urban agglomeration based on weibo check-in big data[J]. Journal of Geo-information Science, 2019, 21(1):68-76. ] |
[1] | 沈泽宇, 丁永生, 孔乔. 降雨径流与洪水淹没模型耦合的应用研究[J]. 地球信息科学学报, 2021, 23(8): 1473-1483. |
[2] | 谢聪慧, 吴世新, 张晨, 孙文涛, 何海芳, 裴韬, 罗格平. 基于谱系聚类的全球各国新冠疫情时间序列特征分析[J]. 地球信息科学学报, 2021, 23(2): 236-245. |
[3] | 巫细波, 赖长强, 葛志专. 政府严控期我国地级市COVID-19疫情的时空集聚、演变及自相关效应研究[J]. 地球信息科学学报, 2021, 23(2): 246-258. |
[4] | 毕佳, 王贤敏, 胡跃译, 罗孟涵, 张俊华, 胡凤昌, 丁子洋. 一种基于改进SEIR模型的突发公共卫生事件风险动态评估与预测方法——以欧洲十国COVID-19为例[J]. 地球信息科学学报, 2021, 23(2): 259-273. |
[5] | 韦原原, 江南, 陈云海, 李响, 杨振凯. 顾及时空对象空间相互作用的疫情风险评估建模与应用[J]. 地球信息科学学报, 2021, 23(2): 274-283. |
[6] | 方云皓, 顾康康. 基于多元数据的中国地理空间疫情风险评估探索——以2020年1月1日至4月11日COVID-19疫情数据为例[J]. 地球信息科学学报, 2021, 23(2): 284-296. |
[7] | 曹中浩, 张健钦, 杨木, 贾礼朋, 邓少存. 基于GIS新冠智能体仿真模型及应用——以广州市为例[J]. 地球信息科学学报, 2021, 23(2): 297-306. |
[8] | 杜毅贤, 徐家鹏, 钟琳颖, 侯盈旭, 沈婕. 网络舆情态势及情感多维特征分析与可视化——以COVID-19疫情为例[J]. 地球信息科学学报, 2021, 23(2): 318-330. |
[9] | 裴韬, 王席, 宋辞, 刘亚溪, 黄强, 舒华, 陈晓, 郭思慧, 周成虎. COVID-19疫情时空分析与建模研究进展[J]. 地球信息科学学报, 2021, 23(2): 188-210. |
[10] | 尹凌, 刘康, 张浩, 奚桂锴, 李璇, 李子垠, 薛建章. 耦合人群移动的COVID-19传染病模型研究进展[J]. 地球信息科学学报, 2021, 23(11): 1894-1909. |
[11] | 崔明洁, 姚霞, 方昊然, 张杨成思, 杨德刚, 裴韬. SARS与COVID-19时空传播差异性及影响因素分析[J]. 地球信息科学学报, 2021, 23(11): 1910-1923. |
[12] | 张浩, 尹凌, 刘康, 毛亮, 冯圣中, 陈洁, 梅树江. 深圳市快速抑制COVID-19疫情的非药物干预措施效果评估:基于智能体的建模研究[J]. 地球信息科学学报, 2021, 23(11): 1936-1945. |
[13] | 刘张, 千家乐, 杜云艳, 王楠, 易嘉伟, 孙晔然, 马廷, 裴韬, 周成虎. 基于多源时空大数据的区际迁徙人群多层次空间分布估算模型——以COVID-19疫情期间自武汉迁出人群为例[J]. 地球信息科学学报, 2020, 22(2): 147-160. |
[14] | 邹雨轩, 吴志峰, 曹峥. 耦合土地利用回归与人口加权模型的PM2.5暴露风险评估[J]. 地球信息科学学报, 2019, 21(7): 1018-1028. |
[15] | 武夕琳, 刘庆生, 刘高焕, 黄翀, 李贺. 高温热浪风险评估研究综述[J]. 地球信息科学学报, 2019, 21(7): 1029-1039. |
|