PREDICTION OF PM<sub>2.5</sub> AND OZONE CONCENTRATION BASED ON VMD-CEEMD DECOMPOSITION AND LSTM

ZHOU Jianguo; WANG Jianyu; WEI Siti

doi:10.13205/j.hjgc.202306021

Volume 41 Issue 6

Jun. 2023

Turn off MathJax

Article Contents

Article Navigation > ENVIRONMENTAL ENGINEERING > 2023 > 41(6): 157-165,221

ZHOU Jianguo, WANG Jianyu, WEI Siti. PREDICTION OF PM2.5 AND OZONE CONCENTRATION BASED ON VMD-CEEMD DECOMPOSITION AND LSTM[J]. ENVIRONMENTAL ENGINEERING , 2023, 41(6): 157-165,221. doi: 10.13205/j.hjgc.202306021

Citation:

ZHOU Jianguo, WANG Jianyu, WEI Siti. PREDICTION OF PM_2.5 AND OZONE CONCENTRATION BASED ON VMD-CEEMD DECOMPOSITION AND LSTM[J]. ENVIRONMENTAL ENGINEERING , 2023, 41(6): 157-165,221. doi: 10.13205/j.hjgc.202306021

Citation:

PDF( 3315 KB)

PREDICTION OF PM_2.5 AND OZONE CONCENTRATION BASED ON VMD-CEEMD DECOMPOSITION AND LSTM

doi: 10.13205/j.hjgc.202306021

School of Business Administration, North China Electric Power University, Baoding 071000, China

Received Date: 2022-09-13
Available Online: 2023-09-02

Abstract

Abstract

Accurate prediction of ozone and PM_2.5 concentration can provide a scientific basis for the prevention and control of photochemical pollution. However, the prediction accuracy of the existing ozone and PM_2.5 concentration prediction models is still not sufficient. Based on the daily average ozone and PM_2.5 concentration data in Nanjing from January 1, 2015, to June 30, 2021, a pollutant concentration prediction model for complementary ensemble empirical mode decomposition (CEEMD) secondary decomposition and long and short-term memory neural network (LSTM) was constructed. Firstly, the ozone and PM_2.5 concentration sequence was decomposed by variational mode decomposition (VMD). Secondly, the CEEMD secondary decomposition was used with residual components, and then all the decomposed subsequences were predicted by LSTM. Finally, the output result was reconstructed to get the final result. The results showed that for the forcast of PM_2.5 and O₃ concentration in Nanjing, comparing with the other models, the model VMD-CEEMD-LSTM proposed in this paper was superior and robust, with the RMSE of ozone and PM_2.5 concentrations of 16.47 and 5.12, respectively. This study could provide valuable references for analyzing ozone and PM_2.5 pollution trend.
- ozone,
- PM_2.5,
- concentration prediction,
- secondary decomposition,
- long- and short-term memory neural networks,
- CEEMD

FullText(HTML)

References(24)

References

[1]	BERO B G, RAZA A, FORSBERG B, et al. Short-term exposure to ozone and mortality in subjects with and without previous cardiovascular disease[J]. Epidemiology. 2016, 27(5):663-669.
[2]	LU X, ZHANG L, WANG X, et al. Rapid increases in warm-season surface ozone and resulting health impact in China since 2013[J]. Environmental Science & Technology Letters. 2020, 7(4):240-247.
[3]	WU W L, XUE W B, ZHENG Y X, et al. Diurnal regulation of VOCs may not be effective in controlling ozone pollution in China[J].Atmospheric Environment, 2021, 256:118442.
[4]	LOAIZA-CEBALLOS M C, MARIN-PALMA D, ZAPATA W, et al. Viral respiratory infections and air pollutants[J]. Air Quality, Atmosphere & Health, 2021,187:109650.
[5]	赵晓东,徐浩然,郭志萍,等.基于区间二型模糊神经网络的臭氧浓度预测[J].计算机应用与软件,2022,39(6):329-335.
[6]	PENDLEBURY D, GRAVEL S, MORAN M D, et al. Impact of chemical lateral boundary conditions in a regional air quality forecast model on surface ozone predictions during stratospheric intrusions[J]. Atmospheric Environment, 2018, 174:148-170.
[7]	PARK S Y, LEE S H, LEE H W. Assimilation of wind profiler observations and its impact on three-dimensional transport of ozone over the Southeast Korean Peninsula[J]. Atmospheric Environment, 2014, 99:660-672.
[8]	PENDLEBURY D, GRAVEL S, MORAN M D, et al. Impact of chemical lateral boundary conditions in a regional air quality forecast model on surface ozone predictions during stratospheric intrusions[J]. Atmospheric Environment, 2018, 174:148-170.
[9]	LUNA A S, PAREDES M L L, DE OLIVEIRA G C G, et al. Prediction of ozone concentration in tropospheric levels using artificial neural networks and support vector machine at Rio de Janeiro, Brazil[J]. Atmospheric Environment, 2014, 98:98-104.
[10]	刘宇轩,应方,叶旭红,等. 基于后向传播神经网络的PM_2.5和臭氧预测研究[J]. 能源工程,2020(5):76-83.
[11]	CHEN S, WANG J, ZHANG H. A hybrid PSO-SVM model based on clustering algorithm for short-term atmospheric pollutant concentration forecasting[J]. Technological Forecasting and Social Change, 2019, 146:41-54.
[12]	董红召,王乐恒,唐伟,等.融合时空特征的PCA-PSO-SVM臭氧(O₃)预测方法研究[J].中国环境科学,2021,41(2):596-605.
[13]	邢红涛,郭江龙,刘书安,等.基于CNN-LSTM混合神经网络模型的NO_<i>x排放预测[J]. 电子测量技术,2022,45(2):98-103.
[14]	WANG L L, LI X, BAI Y L. Short-term wind speed prediction using an extreme learning machine model with error correction[J]. Energy Conversion and Management, 2018, 162:239-250.
[15]	吴子伯,崔云霞,曹炜琦,等.基于CEEMD-BiGRU模型的徐州市大气污染物浓度预测[J].环境工程,2022,40(9):9-18.
[16]	丁子昂,乐曹伟,吴玲玲,等.基于CEEMD-Pearson和深度LSTM混合模型的PM_2.5浓度预测方法[J].计算机科学,2020,47(增刊1):444-449.
[17]	CABANEROS S M, CALAUTIT J K, HUGHES B. Spatial estimation of outdoor NO₂ levels in Central London using deep neural networks and a wavelet decomposition technique[J]. Ecological Modelling, 2020, 424:109017.
[18]	AHANI I K, SALARI M, SHADMAN A. An ensemble multi-step-ahead forecasting system for fine particulate matter in urban areas[J]. Journal of Cleaner Production, 2020, 263:120983.
[19]	ZHU S L, QIU X L, YIN Y R, et al. Two-step-hybrid model based on data preprocessing and intelligent optimization algorithms (CS and GWO) for NO₂ and SO₂ forecasting[J]. Atmospheric Pollution Research, 2019, 10(4):1326-1335.
[20]	HU H L, WANG L, TAO R. Wind speed forecasting based on variational mode decomposition and improved echo state network[J]. Renewable Energy, 2021, 164:729-751.
[21]	SHARMA V, PAREY A. Extraction of weak fault transients using variational mode decomposition for fault diagnosis of gearbox under varying speed[J]. Engineering Failure Analysis, 2020, 107:104204.
[22]	WANG F, YU L, WU A P. Forecasting the electronic waste quantity with a decomposition-ensemble approach[J]. Waste Management, 2021, 120:828-838.
[23]	何哲祥,李雷.一种基于小波变换和LSTM的大气污染物浓度预测模型[J].环境工程,2021,39(3):111-119.
[24]	梁涛,谢高锋,米大斌,等.基于CEEMDAN-SE和LSTM神经网络的PM₁₀浓度预测[J].环境工程,2020,38(2):107-113.