一种面向细粒度空气质量分指数(IAQI)预测的时空因果卷积模型

doi:10.12082/dqxxkx.2023.220321

地球信息科学学报 ›› 2023, Vol. 25 ›› Issue (1): 115-130.doi: 10.12082/dqxxkx.2023.220321

• 地球信息科学理论与方法 • 上一篇下一篇

一种面向细粒度空气质量分指数(IAQI)预测的时空因果卷积模型

张羽民¹(), 赵俊杰¹, 梅强², 刘希亮¹^,^*(), 陈卓栋³, 李建强¹, 王少华⁴, 石宇良¹, 柴金川⁵, 高雨瑶¹, 井小倩¹, 杨念迪¹, 马小焱¹

1.北京工业大学软件学院，北京 100124
2.集美大学航海学院，厦门 361021
3.中国石油审计服务中心北京 100028
4.中国科学院空天信息创新研究院中国科学院数字地球重点实验室，北京 100094
5.中国铁道科学研究院集团有限公司国家铁道试验中心，北京 100015

收稿日期:2022-05-18 修回日期:2022-06-15 出版日期:2023-01-25 发布日期:2023-03-25
通讯作者: *刘希亮（1983—），男，河北衡水人，北京工业大学讲师，研究方向为时空大数据挖掘、物联网、区块链。E-mail: liuxl@bjut.edu.cn
作者简介:张羽民（1997—），女，山西晋中人，硕士生，研究方向为时空大数据挖掘。E-mail: zhangyumin@emails.bjut.edu.cn
基金资助:
国家重点研发计划项目(2020YFB2104400)

A Spatial-temporal Causal Convolution Model for Fine-grained Individual Air Quality Index (IAQI) Prediction

ZHANG Yumin¹(), ZHAO Junjie¹, MEI Qiang², LIU Xiliang¹^,^*(), CHEN Zhuodong³, LI Jianqiang¹, WANG Shaohua⁴, SHI Yuliang¹, CHAI Jinchuan⁵, GAO Yuyao¹, JING Xiaoqian¹, YANG Niandi¹, MA Xiaoyan¹

1. College of Software, Beijing University of Technology, Beijing 100124, China
2. Navigation College, Jimei University, Xiamen 361021, China
3. China National Petroleum Corporation Auditing Service Center, Beijing 100028, China
4. Key Laboratory of Digital Earth Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
5. National Railway Track Test Center, China Academy of Railway Sciences Corporation Limited, Beijing 100015, China

Received:2022-05-18 Revised:2022-06-15 Online:2023-01-25 Published:2023-03-25
Contact: LIU Xiliang
Supported by:
National Key Research and Development Program of China(2020YFB2104400)

摘要/Abstract

摘要：

精确、细粒度空气质量分指数（Individual Air Quality Index, IAQI）预测是空气质量指数（Air Quality Index, AQI）的基础，对于空气质量防治和保护人类身心健康均具有重要意义。目前传统时序建模、循环神经网络（Recurrent Neural Network, RNN）、图卷积网络（Graph Convolutional Network, GCN）等方法难以有效融合时空因素和气象因素，稳定提取监测站点间动态边缘关系。本文提出了基于时空因果卷积网络（Spatial-Temporal Causal Convolution Networks, ST-CCN）的空气质量分指数预测模型ST-CCN-IAQI。首先采用空间注意力机制分析多源空气污染物和气象因素的空间效应；其次利用堆叠膨胀卷积和时间注意力机制提取特征矩阵的时间依赖性特征；最后采用贝叶斯调优方法对膨胀卷积的多种参数进行了调优。本文采用上海市空气监测站空气质量分指数（IAQI-PM2.5）数据展开实验，并采用一系列基线模型（AR、MA、ARMA、ANN、SVR、GRU、LSTM和ST-GCN）与ST-CCN-IAQI效果进行对比。实验结果显示：① 在单测站测试中，ST-CCN-IAQI的RMSE和MAE值分别为9.873、7.469，相比基线模型平均下降了24.95%和16.87%；R²值为0.917，相比基线平均提升了5.69%；② 对全部站点的IAQI-PM2.5、IAQI-PM10和IAQI-NO2的预测，证明了ST-CCN-IAQI具有较强的泛化能力和稳定性。③ 采用Shapley分析方法论证了IAQI-PM10、湿度、IAQI-NO2对IAQI-PM2.5的预测具有较大程度的影响；通过不同数据抽样条件下的Friedman检验，证明了ST-CCN-IAQI对比基线模型有显著的性能提升。ST-CCN-IAQI方法为细粒度IAQI精准预测提供了一种鲁棒可行的解决方案。

关键词: 细粒度空气质量分指数预测, 多源影响因素, 时空注意力, 因果卷积网络, 贝叶斯优化, Shapley分析, Friedman检验, 上海市

Abstract:

Accurate and fine-grained individual Air Quality Index (IAQI) prediction is the basis of Air Quality Index (AQI), which is of great significance for air quality control and human health. Traditional approaches such as time series modeling, Recurrent Neural Network (RNN) or Graph Convolutional Network (GCN) cannot effectively integrate spatial-temporal and meteorological factors and manage dynamic edge relationship among scattered monitoring stations. In this paper, a ST-CCN-IAQI model is proposed based on spatial-temporal causal convolution networks. Firstly, both the spatial effects of multi-source air pollutants and meteorological factors are considered via spatial attention mechanism. Secondly, time-dependent features in causal convolution network are extracted by stacked dilated convolution and time attention. Finally, multiple parameters in ST-CCN-IAQI are tuned by Bayesian optimization. In this paper, the Individual Air Quality Index (IAQI-PM_2.5) data of Shanghai air monitoring station are used to carry out the experiment, and a series of baseline models (AR, MA, ARMA, ANN, SVR, GRU, LSTM, and ST-GCN) are employed to compare with ST-CCN-IAQI. Our results show that: (1) In the single station test, RMSE and MAE values of ST-CCN-IAQI are 9.873 and 7.469, respectively, which decreases by 24.95% and 16.87% on average, respectively; R² is 0.917, about 5.69% higher than that of the baselines; (2) The prediction of IAQI-PM_2.5, IAQI-PM₁₀, and IAQI-NO₂ of all stations proves that ST-CCN-IAQI has strong generalization ability and stability; (3) Shapley analysis shows IAQI-PM₁₀, humidity, and IAQI-NO₂have a great impact on the prediction of IAQI-PM_2.5. Friedman test under different data sampling conditions proves that ST-CCN-IAQI has significant performance improvement by comparisons with baselines. The ST-CCN-IAQI method provides a robust and feasible solution for accurate prediction of fine-grained IAQI.

Key words: individual air quality Index prediction, multi-source factors, temporal and spatial attention, causal convolution network, Bayesian optimization, shapley analysis, Friedman test, Shanghai

张羽民, 赵俊杰, 梅强, 刘希亮, 陈卓栋, 李建强, 王少华, 石宇良, 柴金川, 高雨瑶, 井小倩, 杨念迪, 马小焱. 一种面向细粒度空气质量分指数(IAQI)预测的时空因果卷积模型[J]. 地球信息科学学报, 2023, 25(1): 115-130.DOI:10.12082/dqxxkx.2023.220321

ZHANG Yumin, ZHAO Junjie, MEI Qiang, LIU Xiliang, CHEN Zhuodong, LI Jianqiang, WANG Shaohua, SHI Yuliang, CHAI Jinchuan, GAO Yuyao, JING Xiaoqian, YANG Niandi, MA Xiaoyan. A Spatial-temporal Causal Convolution Model for Fine-grained Individual Air Quality Index (IAQI) Prediction[J]. Journal of Geo-information Science, 2023, 25(1): 115-130.DOI:10.12082/dqxxkx.2023.220321

图/表 19

图1

图2

表1

图3

图4

图5

图6

图7

图8

图9

表2

图10

图11

表3

图12

表4

图13

图14

图15

参考文献 49

[1]	Yang J, Yang Y X, Sun D Q, et al. Influence of urban morphological characteristics on thermal environment[J]. Sustainable Cities and Society, 2021, 72:103045. DOI:10.1016/j.scs.2021.103045 doi: 10.1016/j.scs.2021.103045
[2]	林金煌, 陈文惠, 张岸. 2019年北京市 PM_2.5人群暴露剂量特征分析[J]. 地球信息科学学报, 2020, 22(12):2348-2357. doi: 10.12082/dqxxkx.2020.190624
	[Lin J H, Chen W H, Zhang A. Analysis of PM_2.5 population exposure doses characteristics in Beijing in 2019[J]. Journal of Geo-Information Science, 2020, 22(12):2348-2357. ] DOI:10.12082/dqxxkx.2020.190624 doi: 10.12082/dqxxkx.2020.190624
[3]	Hong C P, Zhang Q, Zhang Y, et al. Impacts of climate change on future air quality and human health in China[J]. PNAS, 2019, 116(35):17193-17200. DOI:10.1073/pnas.1812881116 doi: 10.1073/pnas.1812881116 pmid: 31405979
[4]	Goodkind A L, Tessum C W, Coggins J S, et al. Fine-scale damage estimates of particulate matter air pollution reveal opportunities for location-specific mitigation of emissions[J]. PNAS, 2019, 116(18):8775-8780. DOI:10.1073/pnas.1816102116 doi: 10.1073/pnas.1816102116 pmid: 30962364
[5]	师泽源, 赵俭辉, 张文昱, 等. 基于扩展场强模型的稀疏AQI空间插值新算法[J]. 武汉大学学报·信息科学版, 2017, 42(7):968-974.
	[Shi Z Y, Zhao J H, Zhang W Y, et al. New spatial interpolation algorithm for sparse AQI based on extended field intensity model[J]. Geomatics and Information Science of Wuhan University, 2017, 42(7):968-974. ] DOI:10.13203/j.whugis20140980 doi: 10.13203/j.whugis20140980
[6]	Samal K K R, Panda A K, Babu K S, et al. An improved pollution forecasting model with meteorological impact using multiple imputation and fine-tuning approach[J]. Sustainable Cities and Society, 2021, 70:102923. DOI:10.1016/j.scs.2021.102923 doi: 10.1016/j.scs.2021.102923
[7]	程诗奋, 彭澎, 张恒才, 等. 异质稀疏分布时空数据插值、重构与预测方法探讨[J]. 武汉大学学报·信息科学版, 2020, 45(12):1919-1929.
	[Cheng S F, Peng P, Zhang H C, et al. Review of interpolation, reconstruction and prediction methods for heterogeneous and sparsely distributed geospatial data[J]. Geomatics and Information Science of Wuhan University, 2020, 45(12):1919-1929. ] DOI:10.13203/j.whugis20200488 doi: 10.13203/j.whugis20200488
[8]	冯子钰, 施润和. 中国近地面PM_2.5浓度与排放的时空分布及其关联分析[J]. 地球信息科学学报, 2021, 23(7):1221-1230. doi: 10.12082/dqxxkx.2021.200367
	[Feng Z Y, Shi R H. Spatio-temporal features and the association of ground-level PM_2.5 concentration and its emission in China[J]. Journal of Geo-Information Science, 2021, 23(7):1221-1230. ] DOI:10.12082/dqxxkx.2021.200367 doi: 10.12082/dqxxkx.2021.200367
[9]	Yu B, Yin H T, Zhu Z X. Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting[EB/OL]. 2017: arXiv: 1709.04875. https://arxiv.org/abs/1709.04875
[10]	钟雨桐, 韦晶, 郑月明, 等. 基于静风期污染物的PM_2.5排放清单空间精细化方法[J]. 地球信息科学学报, 2021, 23(11):1971-1983. doi: 10.12082/dqxxkx.2021.210290
	[Zhong Y T, Wei J, Zheng Y M, et al. Spatial refinement method of the PM_2.5 emission inventory based on the quiescent period pollutant[J]. Journal of Geo-Information Science, 2021, 23(11):1971-1983. ] DOI:10.12082/dqxxkx.2021.210290 doi: 10.12082/dqxxkx.2021.210290
[11]	Sharma A, Valdes A C F, Lee Y H. Impact of wildfires on meteorology and air quality (PM_2.5and O3) over western United States during September 2017[J]. Atmosphere, 2022, 13(2):262. DOI:10.3390/atmos13020262 doi: 10.3390/atmos13020262
[12]	Zhao J J, Lin S F, Liu X L, et al. ST-CCN-PM_2.5: Fine-grained PM_2.5 concentration prediction via spatial-temporal causal convolution network[C]// ARIC '21: Proceedings of the 4th ACM SIGSPATIAL International Workshop on Advances in Resilient and Intelligent Cities. 2021:48-55. DOI:10.1145/3486626.3493433 doi: 10.1145/3486626.3493433
[13]	Barthwal A, Acharya D. An IoT based sensing system for modeling and forecasting urban air quality[J]. Wireless Personal Communications, 2021, 116(4):3503-3526. DOI:10.1007/s11277-020-07862-6 doi: 10.1007/s11277-020-07862-6
[14]	Gu Y L, Li B H, Meng Q G. Hybrid interpretable predictive machine learning model for air pollution prediction[J]. Neurocomputing, 2022, 468:123-136. DOI:10.1016/j.neucom.2021.09.051 doi: 10.1016/j.neucom.2021.09.051
[15]	杜震洪, 吴森森, 王中一, 等. 基于地理神经网络加权回归的中国PM_2.5浓度空间分布估算方法[J]. 地球信息科学学报, 2020, 22(1):122-135. doi: 10.12082/dqxxkx.2020.190533
	[Du Z H, Wu S S, Wang Z Y, et al. Estimating ground-level PM_2.5 concentrations across China using geographically neural network weighted regression[J]. Journal of Geo-information Science, 2020, 22(1):122-135. ] DOI:10.12082/dqxxkx.2020.190533 doi: 10.12082/dqxxkx.2020.190533
[16]	李清泉, 高德荃, 杨必胜. 基于模糊支持向量机的城市道路交通状态分类[J]. 吉林大学学报(工学版), 2009, 39(S2):131-134.
	[Li Q Q, Gao D Q, Yang B S. Urban Road traffic status classification based on fuzzy support vector machines[J]. Journal of Jilin University (Engineering and Technology Edition), 2009, 39(S2):131-134. ] DOI:10.1329/j.cnki.jdxbgxb2009.s2.049 doi: 10.1329/j.cnki.jdxbgxb2009.s2.049
[17]	Hung W T, Lu C H S, Alessandrini S, et al. Estimation of PM_2.5 concentrations in New York State: Understanding the influence of vertical mixing on surface PM_2.5 using machine learning[J]. Atmosphere, 2020, 11(12):1303. DOI: 10.3390/atmos11121303 doi: 10.3390/atmos11121303
[18]	Nidzgorska-Lencewicz J. Application of artificial neural networks in the prediction of PM₁₀ levels in the winter months: A case study in the tricity agglomeration, Poland[J]. Atmosphere, 2018, 9(6):203. DOI:10.3390/atmos9060203 doi: 10.3390/atmos9060203
[19]	Motesaddi S, Nowrouz P, Alizadeh B, et al. Sulfur dioxide AQI modeling by artificial neural network in Tehran between 2007 and 2013[J]. Environmental Health Engineering and Management Journal, 2015, 2(4):173-178.
[20]	Bera B, Bhattacharjee S, Sengupta N, et al. PM_2.5 concentration prediction during COVID-19 lockdown over Kolkata metropolitan city, India using MLR and ANN models[J]. Environmental Challenges, 2021, 4:100155. DOI:10.1016/j.envc.2021.100155 doi: 10.1016/j.envc.2021.100155
[21]	Zhou Y L, Chang F J, Chen H, et al. Exploring Copula-based Bayesian Model Averaging with multiple ANNs for PM2. 5 ensemble forecasts[J]. Journal of Cleaner Production, 2020, 263:121528. DOI:10.1016/j.jclepro.2020.121528 doi: 10.1016/j.jclepro.2020.121528
[22]	Hu J L, Yang B, Guo C J, et al. Stochastic origin-destination matrix forecasting using dual-stage graph convolutional, recurrent neural networks[C]// 2020 IEEE 36th International Conference on Data Engineering. IEEE, 2020:1417-1428. DOI:10.1109/ICDE48307.2020.00126 doi: 10.1109/ICDE48307.2020.00126
[23]	Wang J Y, Li X L, Wang T T, et al. A CA-GRU-based model for air quality prediction[J]. International Journal of Ad Hoc and Ubiquitous Computing, 2021, 38(1/2/3):184. DOI:10.1504/ijahuc.2021.119098 doi: 10.1504/ijahuc.2021.119098
[24]	Yang G, Lee H, Lee G. A hybrid deep learning model to forecast particulate matter concentration levels in Seoul, south Korea[J]. Atmosphere, 2020, 11(4):348. DOI:10.3390/atmos11040348 doi: 10.3390/atmos11040348
[25]	Weerakody P B, Wong K W, Wang G J, et al. A review of irregular time series data handling with gated recurrent neural networks[J]. Neurocomputing, 2021, 441:161-178. DOI:10.1016/j.neucom.2021.02.046 doi: 10.1016/j.neucom.2021.02.046
[26]	Li X, Peng L, Yao X J, et al. Long short-term memory neural network for air pollutant concentration predictions: Method development and evaluation[J]. Environmental Pollution (Barking, Essex: 1987), 2017, 231(Pt 1):997-1004. DOI:10.1016/j.envpol.2017.08.114 doi: 10.1016/j.envpol.2017.08.114
[27]	Jiang X C, Luo Y W, Zhang B. Prediction of PM_2.5 concentration based on the LSTM-TSLightGBM variable weight combination model[J]. Atmosphere, 2021, 12(9):1211. DOI:10.3390/atmos12091211 doi: 10.3390/atmos12091211
[28]	Dai H B, Huang G Q, Wang J J, et al. Prediction of air pollutant concentration based on one-dimensional multi-scale CNN-LSTM considering spatial-temporal characteristics: A case study of Xi'an, China[J]. Atmosphere, 2021, 12(12):1626. DOI:10.3390/atmos12121626 doi: 10.3390/atmos12121626
[29]	Park D O, Yoo G W, Park S, et al. Assessment and calibration of a low-cost PM_2.5 sensor using machine learning (HybridLSTM neural network): Feasibility study to build an air quality monitoring system[J]. Atmosphere, 2021, 12(10):1306. DOI:10.3390/atmos12101306 doi: 10.3390/atmos12101306
[30]	Shi L K, Zhang H Z, Xu X, et al. A balanced social LSTM for PM_2.5 concentration prediction based on local spatiotemporal correlation[J]. Chemosphere, 2022, 291:133124. DOI:10.1016/j.chemosphere.2021.133124 doi: 10.1016/j.chemosphere.2021.133124
[31]	Liu B C, Binaykia A, Chang P C, et al. Urban air quality forecasting based on multi-dimensional collaborative Support Vector Regression (SVR): A case study of Beijing-Tianjin-Shijiazhuang[J]. PLoS One, 2017, 12(7):e0179763. DOI:10.1371/journal.pone.0179763 doi: 10.1371/journal.pone.0179763
[32]	Yan R, Liao J Q, Yang J, et al. Multi-hour and multi-site air quality index forecasting in Beijing using CNN, LSTM, CNN-LSTM, and spatiotemporal clustering[J]. Expert Systems With Applications, 2021, 169:114513. DOI:10.1016/j.eswa.2020.114513 doi: 10.1016/j.eswa.2020.114513
[33]	Seo Y, Defferrard M, Vandergheynst P, et al. Structured sequence modeling with graph convolutional recurrent networks[M]// Neural Information Processing. Cham: Springer International Publishing, 2018:362-373. DOI:10.1007/978-3-030-04167-0_33 doi: 10.1007/978-3-030-04167-0_33
[34]	Zhao L, Song Y, Zhang C, et al. T-GCN: A temporal graph convolutional network for traffic prediction[J]. IEEE Transactions on Intelligent Transportation Systems, 2019:21,3848-3858. DOI:10.1109/TITS.2019.2935152 doi: 10.1109/TITS.2019.2935152
[35]	Wu Z H, Pan S R, Long G D, et al. Graph WaveNet for deep spatial-temporal graph modeling[EB/OL]. 2019: arXiv: 1906.00121. https://arxiv.org/abs/1906.00121
[36]	Jiao Y, Wang Z F, Zhang Y. Prediction of air quality index based on LSTM[C]// 2019 IEEE 8th Joint International Information Technology and Artificial Intelligence Conference. IEEE, 2019:17-20. DOI:10.1109/ITAIC.2019.8785602 doi: 10.1109/ITAIC.2019.8785602
[37]	Feng X, Li Q, Zhu Y J, et al. Artificial neural networks forecasting of PM_2.5 pollution using air mass trajectory based geographic model and wavelet transformation[J]. Atmospheric Environment, 2015, 107:118-128. DOI:10.1016/j.atmosenv.2015.02.030 doi: 10.1016/j.atmosenv.2015.02.030
[38]	Pak U, Ma J, Ryu U, et al. Deep learning-based PM_2.5 prediction considering the spatiotemporal correlations: A case study of Beijing, China[J]. The Science of the Total Environment, 2020, 699:133561. DOI:10.1016/j.scitotenv.2019.07.367 doi: 10.1016/j.scitotenv.2019.07.367
[39]	高旺旺, 冯建中, 白林燕, 等. 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. doi: 10.12082/dqxxkx.2020.190360
	[Gao W W, Feng J Z, Bai L Y, et al. Spatiotemporal dynamics and tracing of aerosol over Hainan Island[J]. Journal of Geo-Information Science, 2020, 22(7):1532-1543. ] DOI:10.13203/j.whugis20200488 doi: 10.13203/j.whugis20200488
[40]	Zhang L B, Liu Y Q, Zhao F J. Important meteorological variables for statistical long-term air quality prediction in Eastern China[J]. Theoretical and Applied Climatology, 2018, 134(1/2):25-36. DOI:10.1007/s00704-017-2245-z doi: 10.1007/s00704-017-2245-z
[41]	徐艺武. 北京市PM_2.55污染的时空分布与主要空气污染物的关系[J]. 科技创新与应用, 2016(33):143.
	[Xu Y W. Relationship between the temporal and spatial distribution of PM2.5 pollution and major air pollutants in Beijing[J]. Technology Innovation and Application, 2016(33): 143. ] DOI:CNKI:SUN:CXYY.0.2016-33-123 doi: CNKI:SUN:CXYY.0.2016-33-123
[42]	Zheng Y, Liu F, Hsieh H P. U-air: When urban air quality inference meets big data[C]// Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining, 2013:1436-1444. DOI:10.1145/2487575.2488188 doi: 10.1145/2487575.2488188
[43]	Bai S J, Kolter J Z, Koltun V. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling[J]. arXiv preprint arXiv, 2018,1803.01271. DOI:10.48550/arXiv.1803.01271 doi: 10.48550/arXiv.1803.01271
[44]	Zhang L, Na J M, Zhu J, et al. Spatiotemporal causal convolutional network for forecasting hourly PM_2.5 concentrations in Beijing, China[J]. Computers & Geosciences, 2021, 155:104869. DOI:10.1016/j.cageo.2021.104869 doi: 10.1016/j.cageo.2021.104869
[45]	Mihaljević B, Bielza C, Larrañaga P. Bayesian networks for interpretable machine learning and optimization[J]. Neurocomputing, 2021, 456:648-665. DOI:10.1016/j.neucom.2021.01.138 doi: 10.1016/j.neucom.2021.01.138
[46]	Marcellino M, Stock J H, Watson M W. A comparison of direct and iterated multistep AR methods for forecasting macroeconomic time series[J]. Journal of Econometrics, 2006, 135(1/2):499-526. DOI:10.1016/j.jeconom.2005.07.020 doi: 10.1016/j.jeconom.2005.07.020
[47]	Lento C. Forecasting security returns with simple moving averages[J]. International Business & Economics Research Journal (IBER), 2011, 7(11):11-22. DOI:10.19030/iber.v7i11.3304 doi: 10.19030/iber.v7i11.3304
[48]	Xie Y Y, Zhao B, Zhang L, et al. Spatiotemporal variations of PM_2.5 and PM₁₀ concentrations between 31 Chinese Cities and their relationships with SO₂, NO₂, CO and O3[J]. Particuology, 2015, 20:141-149. DOI:10.1016/j.partic.2015.01.003 doi: 10.1016/j.partic.2015.01.003
[49]	宋明, 韩素芹, 张敏, 等. 天津大气能见度与相对湿度和PM₁₀及 PM_2.5的关系[J]. 气象与环境学报, 2013, 29(2):34-41.
	[Song M, Han S Q, Zhang M, et al. Relationship between visibility and relative humidity, PM₁₀, PM_2.5 in Tianjin[J]. Journal of Meteorology and Environment, 2013, 29(2):34-41. ] DOI:10.3969/j.issn.1673-503X.2013.02.006 doi: 10.3969/j.issn.1673-503X.2013.02.006

数据种类	特征名称	数据类型	单位
空气质量分指数	IAQI-PM_2.5	数值	-
	IAQI-PM₁₀	数值	-
	IAQI-NO₂	数值	-
气象数据	Temperature	数值	℃
	Pressure	数值	hpa
	Humidity	数值	%
	Wind speed	数值	km/h
	Weather	数值	-

模型	RMSE	MAE	R²
AR	12.366	8.397	0.888
MA	17.287	11.897	0.782
ARMA	12.504	8.477	0.885
ANN	13.032	9.030	0.876
SVR	12.821	8.554	0.880
GRU	12.883	8.829	0.878
LSTM	12.853	8.807	0.879
ST-GCN	12.621	8.713	0.884
ST-CCN	9.873	7.469	0.917

站点号	IAQI-PM_2.5			IAQI-PM₁₀			IAQI-NO₂
站点号	RMSE	MAE	R²	RMSE	MAE	R²	RMSE	MAE	R²
1	11.079	8.474	0.885	5.822	4.619	0.812	3.508	2.735	0.872
2	9.873	7.469	0.917	6.447	5.085	0.827	3.378	2.470	0.867
3	10.018	7.766	0.878	5.713	4.524	0.782	2.056	1.567	0.870
4	9.020	6.652	0.909	4.917	3.865	0.869	3.047	2.267	0.859
5	10.293	7.972	0.895	5.904	4.666	0.825	3.070	2.236	0.878
6	9.845	7.611	0.929	5.124	3.961	0.929	3.347	2.593	0.880
7	9.294	7.131	0.930	6.387	4.937	0.778	3.156	2.405	0.870
8	10.004	7.747	0.887	5.011	3.816	0.884	4.181	3.269	0.771
9	9.218	6.926	0.928	7.330	5.826	0.800	2.710	2.029	0.875
平均值	9.849	7.527	0.906	5.850	4.588	0.834	3.161	2.396	0.860

数据集	AR	MA	ARMA	ANN	SVR	GRU	LSTM	ST-GCN	ST-CNN
data_25%	12.35(2)	17.12(9)	12.57(3)	13.03(8)	12.81(5)	12.87(7)	12.84(6)	12.61(4)	9.85(1)
data_50%	12.47(2)	17.67(9)	12.63(3)	13.12(7)	12.92(5)	13.12(7)	13.04(6)	12.76(4)	9.96(1)
data_75%	12.15(2)	17.43(9)	12.45(3)	12.93(7)	12.85(6)	12.97(8)	12.77(5)	12.68(4)	9.62(1)
平均值	2	9	3	7.3	5.3	7.3	5.7	4	1

一种面向细粒度空气质量分指数(IAQI)预测的时空因果卷积模型

A Spatial-temporal Causal Convolution Model for Fine-grained Individual Air Quality Index (IAQI) Prediction

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 19

参考文献 49

相关文章 9

Metrics

本文评价

推荐阅读 0

[1]	黄盛, 李卫江, 朱梦茹, 刘振. 降雨事件对上海地铁通勤客流时空影响的精细尺度研究[J]. 地球信息科学学报, 2022, 24(2): 249-262.
[2]	刘贺, 郭黎, 李豪, 张婉晨, 白翔天. 面实体匹配的集成学习CatBoost方法[J]. 地球信息科学学报, 2022, 24(11): 2198-2211.
[3]	马强, 王亮绪, 龚鑫, 李科. 城市功能区视角下基于POI数据的公共厕所空间布局合理性研究[J]. 地球信息科学学报, 2022, 24(1): 50-62.
[4]	杨练兵, 陈春波, 郑宏伟, 罗格平, 尚白军, Olaf Hellwich. 基于优化随机森林回归模型的土壤盐渍化反演[J]. 地球信息科学学报, 2021, 23(9): 1662-1674.
[5]	冯叶涵, 陈亮, 贺晓冬. 基于百度街景的SVF计算及其在城市热岛研究中的应用[J]. 地球信息科学学报, 2021, 23(11): 1998-2012.
[6]	周琦, 高长春. 城市创意产业空间动态集聚演化的计算与可视优化方法[J]. 地球信息科学学报, 2020, 22(5): 1033-1048.
[7]	陈锐, 陈明剑, 姚翔, 王建光. 利用导航大数据挖掘城市热点区域关联性[J]. 地球信息科学学报, 2019, 21(6): 826-835.
[8]	李明晓, 陈洁, 张恒才, 仇培元, 刘康, 陆锋. 上海市精细时空尺度人口分布估计与特征分析[J]. 地球信息科学学报, 2017, 19(6): 800-807.
[9]	李翔, 陈振杰, 吴洁璇, 汪文祥, 曲乐安, 周琛, 韩肖锋. 基于夜间灯光数据和空间回归模型的城市常住人口格网化方法研究[J]. 地球信息科学学报, 2017, 19(10): 1298-1305.