集成AEC和时空特征的工业园区PM2.5浓度预测

董红召; 廖世凯; 杨强; 应方

PDF(1634 KB)

中国环境科学 ›› 2022, Vol. 42 ›› Issue (10) : 4537-4546.

大气污染与控制

集成AEC和时空特征的工业园区PM_2.5浓度预测

董红召¹, 廖世凯¹, 杨强², 应方³

作者信息 +

PM_2.5 concentration prediction in industrial parks integrating AEC and spatio-temporal characteristics

DONG Hong-zhao¹, LIAO Shi-kai¹, YANG Qiang², YING Fang³

Author information +

文章历史 +

摘要

为实现工业园区企业污染排放精细化管控，捕捉工业园区内企业污染排放与污染物浓度之间的响应关系，提出一种集成大气环境容量（AEC）和时空特征的工业园区PM_2.5浓度预测模型.通过有限体积法获得工业园区日均大气自净能力指数（ASI），结合工业园区日排放数据作为AEC特征；同时利用小波分析和Pearson相关系数法提取时空特征，包括目标监测站PM_2.5浓度的时间变化特征和其与周围监测点PM_2.5的空间相关特征.通过CNN获取训练数据中PM_2.5的关联特征，并利用BILSTM充分反映时间序列训练数据中隐含的关键历史长短期依赖关系，确保快速准确的预测性能，以2018~2020年濮阳市工业园区大气污染物观测数据、气象数据及排放数据进行实验验证.结果表明：本文提出的CNN-BILSTM预测模型相较于传统LSTM模型预测精度提升10%；AEC特征和时空特征有利于提高模型精度和稳定性，集成AEC和时空特征的CNN-BILSTM预测模型在PM_2.5污染天数预测准确率最高，达93%；分季节预测结果表明，秋冬季的预测精度最高.

Abstract

With the purpose of realizing the fine management and control of enterprise pollution emissions and capturing response relationship between enterprise pollution emissions and pollutant concentrations in industrial parks, a PM_2.5 concentration prediction model was proposed to integrate atmospheric environmental capacity (AEC) and spatial-temporal characteristics. Firstly, the daily average atmospheric self-purification capacity index (ASI) was obtained using finite volume method and combined with the daily emission data acted as AEC characteristics. At the same time, temporal characteristics at the target monitoring station and spatial characteristics including its surrounding monitoring points of PM_2.5concentration were showed by wavelet analysis and Pearson correlation coefficient method. Then, in order to guarantee the fast and accurate prediction performance, the correlation characteristics of PM_2.5 in training data were obtained by CNN, and BILSTM was used to fully reflect the key historical long short-term dependencies implied in time series training data. The air pollutant observation data, meteorological data and emission data of Puyang Industrial Park from 2018 to 2020 were applied to experimental verification. The results show that the CNN-BILSTM prediction model proposed in this paper improves the prediction accuracy by 10% compared with the traditional LSTM model. AEC and spatio-temporal features are well situated to improve accuracy and stability of model. The CNN-BILSTM prediction model integrating AEC and spatio-temporal features has the highest prediction accuracy during PM_2.5 pollution period, which is up to 93%. Moreover, the seasonal prediction results performed the highest prediction accuracy in autumn and winter.

导出引用

董红召, 廖世凯, 杨强, 应方. 集成AEC和时空特征的工业园区PM_2.5浓度预测[J]. 中国环境科学. 2022, 42(10): 4537-4546

DONG Hong-zhao, LIAO Shi-kai, YANG Qiang, YING Fang. PM_2.5 concentration prediction in industrial parks integrating AEC and spatio-temporal characteristics[J]. China Environmental Science. 2022, 42(10): 4537-4546

中图分类号： X513

参考文献

[1] Lu D, Xu J, Yang D, et al.Spatio-temporal variation and influence factors of PM_2.5 concentrations in China from 1998 to 2014[J].Atmospheric Pollution Research, 2017,8(6):1151-1159.
[2] 王郁,徐大海,孙俊英.新疆地区大气环境容量系数的气候特征及其在空气质量变化中的作用[J].环境科学学报, 2021,41(12):5073-5082.Wang Y, Xu D H, Sun J Y.Climatic characteristics of atmospheric environmental capacity coefficient in Xinjiang and its role in air quality changes[J].Acta Scientiae Circumstantiae, 2021,41(12):5073-5082.
[3] 朱蓉,张存杰,梅梅.大气自净能力指数的气候特征与应用研究[J].中国环境科学, 2018,38(10):3601-3610.Zhu R, Zhang C J, Mei M.Research on climate characteristics and application of atmospheric self-purification ability index[J].China Environmental Science, 2018,38(10):3601-3610.
[4] 徐大海.大气环境容量系数A值频率曲线拟合及其应用[J].中国环境科学, 2016,36(10):2913-2922.Xu D H.Frequency curve fitting and application of atmospheric environment capacity coefficient A value[J].China Environmental Science, 2016,36(10):2913-2922.
[5] 孙宝磊,孙暠,张朝能,等.基于BP神经网络的大气污染物浓度预测[J].环境科学学报, 2017,37(5):1864-1871.Sun B L, Sun H, Zhang C N, et al.Prediction of atmospheric pollutant concentration based on BP neural network[J].Acta Scientiae Circumstantiae, 2017,37(5):1864-1871.
[6] Dennis R L, Byun D W, Novak J H.The next generation of integrated air quality modeling:EPA's models-3[J].Atmospheric Environment, 1996,30(12):1925-1938.
[7] 周广强,谢英,吴剑斌,等.基于WRF-Chem模式的华东区域PM_2.5预报及偏差原因[J].中国环境科学, 2016,36(8):2251-2259.Zhou G Q, Xie Y, Wu J B, et al.WRF-Chem based PM_2.5 forecast and bias analysis over the East China Region[J].China Environmental Science, 2016,36(8):2251-2259.
[8] Qingxin W, Qiaolin Z, Jinhua T, et al.Estimating PM_2.5 concentrations based on MODIS AOD and NAQPMS data over Beijing Tianjin Hebei.[J].Sensors, 2019,19(5):1207.
[9] Sun W, Sun J.Daily PM_2.5 concentration prediction based on principal component analysis and LSSVM optimized by cuckoo search algorithm[J].Journal of Environmental Management, 2017,188:144-152.
[10] 宋国君,国潇丹,杨啸,等.沈阳市PM_2.5浓度ARIMA-SVM组合预测研究[J].中国环境科学, 2018,38(11):4031-4039.Song G J, Guo X D, Yang X, et al.ARIMA-SVM combination prediction of PM_2.5 concentration in Shenyang[J].China Environmental Science, 2018,38(11):4031-4039.
[11] 李建新,刘小生,刘静,等.基于MRMR-HK-SVM模型的PM_2.5浓度预测[J].中国环境科学, 2019,39(6):2304-2310.Li J X, Liu X S, Liu J, et al.Prediction of PM_2.5 concentration based on MRMR-HK-SVM model[J].China Environmental Science, 2019, 39(6):2304-2310.
[12] 侯俊雄,李琦,朱亚杰,等.基于随机森林的PM_2.5实时预报系统[J].测绘科学, 2017,42(1):1-6.Hou J X, Li Q, Zhu Y J, et al.Real-time forecasting system of PM_2.5 concentration based on spark framework and random forest model[J].Science of Surveying and Mapping, 2017,42(1):1-6.
[13] 任才溶,谢刚.基于随机森林和气象参数的PM_2.5浓度等级预测[J].计算机工程与应用, 2019,55(2):213-220.Ren C R, Xie G.Prediction of PM_2.5 concentration level based on random forest and meteorological parameters[J].Computer Engineering and Applications, 2019,55(2):213-220.
[14] 夏晓圣,陈菁菁,王佳佳,等.基于随机森林模型的中国PM_2.5浓度影响因素分析[J].环境科学, 2020,41(5):2057-2065.Xia X S, Chen J J, Wang J J, et al.PM_2.5 concentration influencing factors in China based on the random forest model[J].Environmental Science, 2020,41(5):2057-2065.
[15] 王敏,邹滨,郭宇,等.基于BP人工神经网络的城市PM_2.5浓度空间预测[J].环境污染与防治, 2013,35(9):63-66.Wang M, Zhou B, Guo Y, et al.BP artificial neural network-based analysis of spatial variability of urban PM_2.5 concentration[J].Environmental Pollution and Control, 2013,35(9):63-66.
[16] Dai H, Huang G, Zeng H, et al.PM_2.5 concentration prediction based on spatiotemporal feature selection using XGBoost-MSCNN-GA-LSTM[J].Sustainability, 2021,13(21):12071.
[17] Li S, Xie G, Ren J, et al.Urban PM_2.5 concentration prediction via attention-based CNN-LSTM[J].Applied Sciences, 2020,10(6):1953.
[18] 邓丽,邬群勇,杨水荣.融合SSAE深度特征学习和LSTM网络的PM_2.5小时浓度预测[J].环境科学学报, 2020,40(9):3422-3434.Deng L, Wu Q Y, Yang S R.Prediction of PM_2.5 hourly concentration using SSAE deep feature learning and LSTM network[J].Acta Scientiae Circumstantiae, 2020,40(9):3422-3434.
[19] Liao Q, Zhu M, Wu L, et al.Deep learning for air quality forecasts:a review[J].Current Pollution Reports, 2020,6(4):399-409.
[20] 武艳,杜景林,全力,等.一种基于PCA-IPSO的BiLSTM神经网络预测PM[J].现代电子技术, 2022,45(1):119-124.Wu Y, Du J L, Quan L, et al.A BiLSTM neural network based on PCA-IPSO to predict PM[J].Modern Electronic Technology, 2022,45(1):119-124.
[21] Zhang M, Wu D, Xue R.Hourly prediction of PM_2.5 concentration in Beijing based on Bi-LSTM neural network[J].Multimedia Tools and Applications, 2021,80(16):24455-24468.
[22] Liu H, Dong S.A novel hybrid ensemble model for hourly PM_2.5 forecasting using multiple neural networks:a case study in China[J].Air Quality, Atmosphere & amp:Health, 2020,13(12):1411-1420.
[23] 康俊锋,谭建林,方雷,等.XGBoost-LSTM变权组合模型支持下短期PM_2.5浓度预测——以上海为例[J].中国环境科学, 2021,41(9):4016-4025.Kang J F, Tan J L, Fang L, et al.Short-term PM_2.5 concentration prediction under the support of XGBoost-LSTM variable weight combination model:Taking Shanghai as an example[J].China Environmental Science, 2021,41(9):4016-4025.
[24] 梁泽,王玥瑶,岳远紊,等.耦合遗传算法与RBF神经网络的PM_2.5浓度预测模型[J].中国环境科学, 2020,40(2):523-529.Liang Z, Wang Y Y, Yue Y X, et al.PM_2.5 concentration prediction model coupled with genetic algorithm and RBF neural network[J].China Environmental Science, 2020,40(2):523-529.
[25] 董红召,王乐恒,唐伟,等.集成时空特征的PCA-PSO-SVM臭氧(O₃)预测方法研究[J].中国环境科学, 2021,41(2):596-605.Dong H Z, Wang L H, Tang W, et al.Research on PCA-PSO-SVM ozone (O₃) prediction method integrating spatiotemporal features[J].China Environmental Science, 2021,41(2):596-605.
[26] 王平,张红,秦作栋,等.基于wavelet-SVM的PM₁₀浓度时序数据预测[J].环境科学, 2017,38(8):3153-3161.Wang P, Zhong H, Qin Z D, et al.Prediction of PM₁₀ concentration time series data based on wavelet-SVM[J].Environmental Science, 2017,38(8):3153-3161.
[27] 杨玉璟,李彬,李源昊.濮阳市西部工业园区对城区大气环境影响分析[J].濮阳职业技术学院学报, 2015,28(4):155-157.Yang Y J, Li B, Li Y H.Analysis of the impact of the western industrial park in Puyang on the atmospheric environment in the urban area[J].Journal of Puyang Vocational and Technical College, 2015,28(4):155-157.
[28] 濮阳新闻网.市环境污染防治攻坚办关于工业园区管委会对工[EB/OL].http://www.xiaoyukun.com/a/puyangxinwen/20191007/4019.html2019-09-28/2019-10-07.Puyang News Network.The Municipal Environmental Pollution Prevention and Control Office on the Industrial Park Management Committee's work[EB/OL].http://www.xiaoyukun.com/a/puyangxinwen/20191007/4019.html2019-09-28/2019-10-07.
[29] 中华人民共和国生态环境部.濮阳市工业园区督办问题清单[EB/OL].https://www.mee.gov.cn/xxgk2018/xxgk/xxgk12/201811/W020181106387569066991.pdf.Ministry of Ecology and Environment of the People's Republic of China.List of Issues for Supervision of Puyang Industrial Park[EB/OL].https://www.mee.gov.cn/xxgk2018/xxgk/xxgk12/201811/W020181106387569066991.pdf.
[30] 曾昭亮,郭建平,马大喜.基于江西地区多卫星数据的气溶胶立体分布研究[J].大气与环境光学学报, 2016,11(5):75-84.Zeng Z L, Guo J P, Ma D X.Research of aerosol three-dimensional distribution based on multi-satellite data over Jiangxi[J].Journal of Atmospheric and Environmental Optics, 2016,11(5):75-84.
[31] 中国天气网.濮阳[EB/OL].www.weather.com.cn/html/2010-05-10.China Weather Network.Puyang[EB/OL].www.weather.com.cn/html/2010-05-10.
[32] 侯俊雄,李琦,朱亚杰,等.基于随机森林的PM_2.5实时预报系统[J].测绘科学, 2017,42(1):1-6.Hou J X, Li Q, Zhu Y J, et al.Real-time forecasting system of PM_2.5 concentration based on spark framework and random forest model[J].Science of Surveying and Mapping, 2017,42(1):1-6.
[33] 赵文芳,林润生,唐伟,等.基于深度学习的PM_2.5短期预测模型[J].南京师大学报:自然科学版, 2019,42(3):32-41.Zhao W F, Lin R S, Tang W, et al.PM_2.5 short-term prediction model based on deep learning[J].Journal of Nanjing Normal University (Natural Science Edition), 2019,42(3):32-41.
[34] Yang B, Liu R, Zio E.Remaining useful life prediction based on a double-convolutional neural network architecture[J].IEEE Transactions on Industrial Electronics, 2019,66(12):9521-9530.
[35] Liu R, Yang B, Hauptmann A G.Simultaneous bearing fault recognition and remaining useful life prediction using joint-loss convolutional neural network[J].IEEE Transactions on industrial informatics, 2019,16(1):87-96.
[36] Pa A, Hk B, Bkp A.Prediction, and analysis of COVID-19positive cases using deep learning models:A descriptive case study of India[J].Chaos, Solitons & Fractals, 2020,139:110017.
[37] Shahid F, Zameer A, Muneeb M.Predictions for COVID-19 with deep learning models of LSTM, GRU and Bi-LSTM[J].Chaos Solitons & Fractals, 2020,140:110212.
[38] HJ 633-2012环境空气质量指数(AQI)技术规定(试行)[S].HJ 633-2012 Technical specification of ambient air quality index (AQI) (Trial)[S].