An Efficient Method for Capturing the High Peak Concentrations of PM<sub>2.5</sub> Using Gaussian-Filtered Deep Learning
This paper proposes a deep learning model that integrates a convolutional neural network with a gate circulation unit that captures patterns of high-peak PM<sub>2.5</sub> concentrations. The purpose is to accurately predict high-peak PM<sub>2.5</sub> concentration data that c...
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | article |
| Lenguaje: | EN |
| Publicado: |
MDPI AG
2021
|
| Materias: | |
| Acceso en línea: | https://doaj.org/article/c0f60351b6164ad097f32db20f70547e |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| id |
oai:doaj.org-article:c0f60351b6164ad097f32db20f70547e |
|---|---|
| record_format |
dspace |
| spelling |
oai:doaj.org-article:c0f60351b6164ad097f32db20f70547e2021-11-11T19:34:59ZAn Efficient Method for Capturing the High Peak Concentrations of PM<sub>2.5</sub> Using Gaussian-Filtered Deep Learning10.3390/su1321118892071-1050https://doaj.org/article/c0f60351b6164ad097f32db20f70547e2021-10-01T00:00:00Zhttps://www.mdpi.com/2071-1050/13/21/11889https://doaj.org/toc/2071-1050This paper proposes a deep learning model that integrates a convolutional neural network with a gate circulation unit that captures patterns of high-peak PM<sub>2.5</sub> concentrations. The purpose is to accurately predict high-peak PM<sub>2.5</sub> concentration data that cannot be trained in general deep learning models. For the training of the proposed model, we used all available weather and air quality data for three years from 2015 to 2017 from 25 stations of the National Institute of Environmental Research (NIER) and the Korea Meteorological Administration (KMA) observatory in Seoul, South Korea. Our model trained three years of data and predicted high-peak PM<sub>2.5</sub> concentrations for the year 2018. In addition, we propose a Gaussian filter algorithm as a preprocessing method for capturing high concentrations of PM<sub>2.5</sub> in the Seoul area and predicting them more accurately. This model overcomes the limitations of conventional deep learning approaches that are unable to predict high peak PM<sub>2.5</sub> concentrations. Comparing model measurements at each of the 25 monitoring sites in 2018, we found that the deep learning model with a Gaussian filter achieved an index of agreement of 0.73–0.89 and a proportion of correctness of 0.89–0.96, and compared to the conventional deep learning method (average POC = 0.85), the Gaussian filter algorithm (average POC = 0.94) improved the accuracy of high-concentration PM<sub>2.5</sub> prediction by an average of about 9%. Applying this algorithm in the preprocessing stage could be updated to predict the risk of high PM<sub>2.5</sub> concentrations in real time.Inchoon YeoYunsoo ChoiMDPI AGarticleair-qualityGaussian filteringdeep learningCNNhigh peak forecasting of PM<sub>2.5</sub>Environmental effects of industries and plantsTD194-195Renewable energy sourcesTJ807-830Environmental sciencesGE1-350ENSustainability, Vol 13, Iss 11889, p 11889 (2021) |
| institution |
DOAJ |
| collection |
DOAJ |
| language |
EN |
| topic |
air-quality Gaussian filtering deep learning CNN high peak forecasting of PM<sub>2.5</sub> Environmental effects of industries and plants TD194-195 Renewable energy sources TJ807-830 Environmental sciences GE1-350 |
| spellingShingle |
air-quality Gaussian filtering deep learning CNN high peak forecasting of PM<sub>2.5</sub> Environmental effects of industries and plants TD194-195 Renewable energy sources TJ807-830 Environmental sciences GE1-350 Inchoon Yeo Yunsoo Choi An Efficient Method for Capturing the High Peak Concentrations of PM<sub>2.5</sub> Using Gaussian-Filtered Deep Learning |
| description |
This paper proposes a deep learning model that integrates a convolutional neural network with a gate circulation unit that captures patterns of high-peak PM<sub>2.5</sub> concentrations. The purpose is to accurately predict high-peak PM<sub>2.5</sub> concentration data that cannot be trained in general deep learning models. For the training of the proposed model, we used all available weather and air quality data for three years from 2015 to 2017 from 25 stations of the National Institute of Environmental Research (NIER) and the Korea Meteorological Administration (KMA) observatory in Seoul, South Korea. Our model trained three years of data and predicted high-peak PM<sub>2.5</sub> concentrations for the year 2018. In addition, we propose a Gaussian filter algorithm as a preprocessing method for capturing high concentrations of PM<sub>2.5</sub> in the Seoul area and predicting them more accurately. This model overcomes the limitations of conventional deep learning approaches that are unable to predict high peak PM<sub>2.5</sub> concentrations. Comparing model measurements at each of the 25 monitoring sites in 2018, we found that the deep learning model with a Gaussian filter achieved an index of agreement of 0.73–0.89 and a proportion of correctness of 0.89–0.96, and compared to the conventional deep learning method (average POC = 0.85), the Gaussian filter algorithm (average POC = 0.94) improved the accuracy of high-concentration PM<sub>2.5</sub> prediction by an average of about 9%. Applying this algorithm in the preprocessing stage could be updated to predict the risk of high PM<sub>2.5</sub> concentrations in real time. |
| format |
article |
| author |
Inchoon Yeo Yunsoo Choi |
| author_facet |
Inchoon Yeo Yunsoo Choi |
| author_sort |
Inchoon Yeo |
| title |
An Efficient Method for Capturing the High Peak Concentrations of PM<sub>2.5</sub> Using Gaussian-Filtered Deep Learning |
| title_short |
An Efficient Method for Capturing the High Peak Concentrations of PM<sub>2.5</sub> Using Gaussian-Filtered Deep Learning |
| title_full |
An Efficient Method for Capturing the High Peak Concentrations of PM<sub>2.5</sub> Using Gaussian-Filtered Deep Learning |
| title_fullStr |
An Efficient Method for Capturing the High Peak Concentrations of PM<sub>2.5</sub> Using Gaussian-Filtered Deep Learning |
| title_full_unstemmed |
An Efficient Method for Capturing the High Peak Concentrations of PM<sub>2.5</sub> Using Gaussian-Filtered Deep Learning |
| title_sort |
efficient method for capturing the high peak concentrations of pm<sub>2.5</sub> using gaussian-filtered deep learning |
| publisher |
MDPI AG |
| publishDate |
2021 |
| url |
https://doaj.org/article/c0f60351b6164ad097f32db20f70547e |
| work_keys_str_mv |
AT inchoonyeo anefficientmethodforcapturingthehighpeakconcentrationsofpmsub25subusinggaussianfiltereddeeplearning AT yunsoochoi anefficientmethodforcapturingthehighpeakconcentrationsofpmsub25subusinggaussianfiltereddeeplearning AT inchoonyeo efficientmethodforcapturingthehighpeakconcentrationsofpmsub25subusinggaussianfiltereddeeplearning AT yunsoochoi efficientmethodforcapturingthehighpeakconcentrationsofpmsub25subusinggaussianfiltereddeeplearning |
| _version_ |
1718431490513043456 |