Combustion Instability Monitoring through Deep-Learning-Based Classification of Sequential High-Speed Flame Images
In this study, novel deep learning models based on high-speed flame images are proposed to diagnose the combustion instability of a gas turbine. Two different network layers that can be combined with any existing backbone network are established—(1) An early-fusion layer that can learn to extract th...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-05-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/9/5/848 |
| _version_ | 1797567466045440000 |
|---|---|
| author | Ouk Choi Jongwun Choi Namkeun Kim Min Chul Lee |
| author_facet | Ouk Choi Jongwun Choi Namkeun Kim Min Chul Lee |
| author_sort | Ouk Choi |
| collection | DOAJ |
| description | In this study, novel deep learning models based on high-speed flame images are proposed to diagnose the combustion instability of a gas turbine. Two different network layers that can be combined with any existing backbone network are established—(1) An early-fusion layer that can learn to extract the power spectral density of subsequent image frames, which is time-invariant under certain conditions. (2) A late-fusion layer which combines the outputs of a backbone network at different time steps to predict the current combustion state. The performance of the proposed models is validated by the dataset of high speed flame images, which have been obtained in a gas turbine combustor during the transient process from stable condition to unstable condition and vice versa. Excellent performance is achieved for all test cases with high accuracy of 95.1–98.6% and a short processing time of 5.2–12.2 ms. Interestingly, simply increasing the number of input images is as competitive as combining the proposed early-fusion layer to a backbone network. In addition, using handcrafted weights for the late-fusion layer is shown to be more effective than using learned weights. From the results, the best combination is selected as the ResNet-18 model combined with our proposed fusion layers over 16 time-steps. The proposed deep learning method is proven as a potential tool for combustion instability identification and expected to be a promising tool for combustion instability prediction as well. |
| first_indexed | 2024-03-10T19:43:21Z |
| format | Article |
| id | doaj.art-f500e23986a5496a8cca278ca1085cb3 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-10T19:43:21Z |
| publishDate | 2020-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-f500e23986a5496a8cca278ca1085cb32023-11-20T01:06:10ZengMDPI AGElectronics2079-92922020-05-019584810.3390/electronics9050848Combustion Instability Monitoring through Deep-Learning-Based Classification of Sequential High-Speed Flame ImagesOuk Choi0Jongwun Choi1Namkeun Kim2Min Chul Lee3Department of Electronics Engineering, Incheon National University, Incheon 22012, KoreaDepartment of Mechanical Engineering, Incheon National University, Incheon 22012, KoreaDepartment of Mechanical Engineering, Incheon National University, Incheon 22012, KoreaDepartment of Safety Engineering, Incheon National University, Incheon 22012, KoreaIn this study, novel deep learning models based on high-speed flame images are proposed to diagnose the combustion instability of a gas turbine. Two different network layers that can be combined with any existing backbone network are established—(1) An early-fusion layer that can learn to extract the power spectral density of subsequent image frames, which is time-invariant under certain conditions. (2) A late-fusion layer which combines the outputs of a backbone network at different time steps to predict the current combustion state. The performance of the proposed models is validated by the dataset of high speed flame images, which have been obtained in a gas turbine combustor during the transient process from stable condition to unstable condition and vice versa. Excellent performance is achieved for all test cases with high accuracy of 95.1–98.6% and a short processing time of 5.2–12.2 ms. Interestingly, simply increasing the number of input images is as competitive as combining the proposed early-fusion layer to a backbone network. In addition, using handcrafted weights for the late-fusion layer is shown to be more effective than using learned weights. From the results, the best combination is selected as the ResNet-18 model combined with our proposed fusion layers over 16 time-steps. The proposed deep learning method is proven as a potential tool for combustion instability identification and expected to be a promising tool for combustion instability prediction as well.https://www.mdpi.com/2079-9292/9/5/848combustion instabilityflame imagingdeep learningresidual networkpower spectral densitytemporal smoothing |
| spellingShingle | Ouk Choi Jongwun Choi Namkeun Kim Min Chul Lee Combustion Instability Monitoring through Deep-Learning-Based Classification of Sequential High-Speed Flame Images Electronics combustion instability flame imaging deep learning residual network power spectral density temporal smoothing |
| title | Combustion Instability Monitoring through Deep-Learning-Based Classification of Sequential High-Speed Flame Images |
| title_full | Combustion Instability Monitoring through Deep-Learning-Based Classification of Sequential High-Speed Flame Images |
| title_fullStr | Combustion Instability Monitoring through Deep-Learning-Based Classification of Sequential High-Speed Flame Images |
| title_full_unstemmed | Combustion Instability Monitoring through Deep-Learning-Based Classification of Sequential High-Speed Flame Images |
| title_short | Combustion Instability Monitoring through Deep-Learning-Based Classification of Sequential High-Speed Flame Images |
| title_sort | combustion instability monitoring through deep learning based classification of sequential high speed flame images |
| topic | combustion instability flame imaging deep learning residual network power spectral density temporal smoothing |
| url | https://www.mdpi.com/2079-9292/9/5/848 |
| work_keys_str_mv | AT oukchoi combustioninstabilitymonitoringthroughdeeplearningbasedclassificationofsequentialhighspeedflameimages AT jongwunchoi combustioninstabilitymonitoringthroughdeeplearningbasedclassificationofsequentialhighspeedflameimages AT namkeunkim combustioninstabilitymonitoringthroughdeeplearningbasedclassificationofsequentialhighspeedflameimages AT minchullee combustioninstabilitymonitoringthroughdeeplearningbasedclassificationofsequentialhighspeedflameimages |