Improving Intra-Urban Prediction of Atmospheric Fine Particles Using a Hybrid Deep Learning Approach

被引:2
|
作者
Zhang, Zhengyu [1 ]
Ren, Jiuchun [1 ]
Chang, Yunhua [2 ]
机构
[1] Fudan Univ, Sch Informat Sci & Engn, Dept Commun Engn, Shanghai 200433, Peoples R China
[2] Nanjing Univ Informat Sci & Technol, Collaborat Innovat Ctr Forecast & Evaluat Meteorol, NUIST Ctr Atmospher Environm, Nanjing 210044, Peoples R China
基金
中国国家自然科学基金;
关键词
fine particles; deep learning; hybrid model; street-level pollution; PM2; 5; prediction; NETWORKS; EXPOSURE;
D O I
10.3390/atmos14030599
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Growing evidence links intra-urban gradients in atmospheric fine particles (PM2.5), a complex and variable cocktail of toxic chemicals, to adverse health outcomes. Here, we propose an improved hierarchical deep learning model framework to estimate the hourly variation of PM2.5 mass concentration at the street level. By using a full-year monitoring data (including meteorological parameters, hourly concentrations of PM2.5, and gaseous precursors) from multiple stations in Shanghai, the largest city in China, as a training dataset, we first apply a convolutional neural network to obtain cross-domain and time-series features so that the inherent features of air quality and meteorological data associated with PM2.5 can be effectively extracted. Next, a Gaussian weight calculation layer is used to determine the potential interaction effects between different regions and neighboring stations. Finally, a long and short-term memory model layer is used to efficiently extract the temporal evolution characteristics of PM2.5 concentrations from the previous output layer. Further comparative analysis reveals that our proposed model framework significantly outperforms previous benchmark methods in terms of the stability and accuracy of PM2.5 prediction, which has important implications for the intra-urban health assessment of PM2.5-related pollution exposures.
引用
收藏
页数:15
相关论文
共 50 条
  • [41] A hybrid deep learning and handcrafted feature approach for the prediction of protein structural class
    Yagoubi, Rached
    Moussaoui, Abdelouahab
    Yagoubi, Mohamed Bachir
    COMPUTER SCIENCE JOURNAL OF MOLDOVA, 2022, 30 (01) : 93 - 108
  • [42] A hybrid attention-based deep learning approach for wind power prediction
    Ma, Zhengjing
    Mei, Gang
    APPLIED ENERGY, 2022, 323
  • [43] Hybrid deep learning approach for rock tunnel deformation prediction based on spatio-temporal patterns
    Sun, Junfeng
    Fang, Yong
    Luo, Hu
    Yao, Zhigang
    Xiang, Long
    Wang, Jianfeng
    Wang, Yubo
    Jiang, Yifan
    UNDERGROUND SPACE, 2025, 20 : 100 - 118
  • [44] Roll motion prediction using a hybrid deep learning and ARIMA model
    Suhermi, Novri
    Suhartono
    Prastyo, Dedy Dwi
    Ali, Baharuddin
    INNS CONFERENCE ON BIG DATA AND DEEP LEARNING, 2018, 144 : 251 - 258
  • [45] Deep learning-based feature selection and prediction system for autism spectrum disorder using a hybrid meta-heuristics approach
    Raja, K. Chola
    Kannimuthu, S.
    JOURNAL OF INTELLIGENT & FUZZY SYSTEMS, 2023, 45 (01) : 797 - 807
  • [46] Automatic Malignant and Benign Skin Cancer Classification Using a Hybrid Deep Learning Approach
    Bassel, Atheer
    Abdulkareem, Amjed Basil
    Alyasseri, Zaid Abdi Alkareem
    Sani, Nor Samsiah
    Mohammed, Husam Jasim
    DIAGNOSTICS, 2022, 12 (10)
  • [47] A New Approach to Load Shedding Prediction in GECOL Using Deep Learning Neural Network
    Abusida, Ashraf Mohammed
    Hancerliogullari, Aybaba
    INTERNATIONAL JOURNAL OF COMPUTER SCIENCE AND NETWORK SECURITY, 2022, 22 (03): : 220 - 228
  • [48] An integrated feature learning approach using deep learning for travel time prediction
    Abdollahi, Mohammad
    Khaleghi, Tannaz
    Yang, Kai
    EXPERT SYSTEMS WITH APPLICATIONS, 2020, 139
  • [49] Air quality prediction using CNN+LSTM-based hybrid deep learning architecture
    Aysenur Gilik
    Arif Selcuk Ogrenci
    Atilla Ozmen
    Environmental Science and Pollution Research, 2022, 29 : 11920 - 11938
  • [50] A hybrid attention-based deep learning approach for wind power prediction
    Ma, Zhengjing
    Mei, Gang
    APPLIED ENERGY, 2022, 323