Resource-Efficient Acoustic Full-Waveform Inversion via Dual-Branch Physics-Informed RNN with Scale Decomposition
Full-waveform velocity inversion has long been a primary focus in seismic exploration. Full-waveform inversion techniques employing physics-informed recurrent neural networks (PIRNNs) have recently gained significant scholarly attention. However, these approaches demand considerable storage to captu...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/15/2/941 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832589204995964928 |
|---|---|
| author | Cai Lu Jijun Liu Liyuan Qu Jianbo Gao Hanpeng Cai Jiandong Liang |
| author_facet | Cai Lu Jijun Liu Liyuan Qu Jianbo Gao Hanpeng Cai Jiandong Liang |
| author_sort | Cai Lu |
| collection | DOAJ |
| description | Full-waveform velocity inversion has long been a primary focus in seismic exploration. Full-waveform inversion techniques employing physics-informed recurrent neural networks (PIRNNs) have recently gained significant scholarly attention. However, these approaches demand considerable storage to capture spatiotemporal seismic wave propagation fields and their gradient information, often exceeding the memory capabilities of current GPU resources during field data processing. This study proposes a full-waveform inversion method utilizing a dual-branch PIRNN architecture to effectively minimize GPU resource consumption. The primary PIRNN branch performs forward-wave equation modeling at the original scale and computes the discrepancy between synthetic and observed seismic records. Additionally, a downscaled spatiotemporal PIRNN branch is introduced, transforming the original-scale error into a loss function via scale decomposition, which drives the inversion process in the downscaled domain. This dual-branch design necessitates recording only the spatiotemporal field and gradient information of the downscaled branch, significantly reducing GPU memory requirements. The proposed dual-branch PIRNN framework was validated through full-waveform inversions on synthetic horizontal-layer models and the Marmousi model across various scales. The results demonstrate that this approach markedly reduces resource consumption while maintaining high inversion accuracy. |
| format | Article |
| id | doaj-art-a194dce7243d43509889afb377d580f5 |
| institution | Kabale University |
| issn | 2076-3417 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj-art-a194dce7243d43509889afb377d580f52025-01-24T13:21:26ZengMDPI AGApplied Sciences2076-34172025-01-0115294110.3390/app15020941Resource-Efficient Acoustic Full-Waveform Inversion via Dual-Branch Physics-Informed RNN with Scale DecompositionCai Lu0Jijun Liu1Liyuan Qu2Jianbo Gao3Hanpeng Cai4Jiandong Liang5The School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu 611731, ChinaThe School of Resources and Environment, University of Electronic Science and Technology of China, Chengdu 611731, ChinaThe School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu 611731, ChinaThe School of Resources and Environment, University of Electronic Science and Technology of China, Chengdu 611731, ChinaThe School of Resources and Environment, University of Electronic Science and Technology of China, Chengdu 611731, ChinaThe School of Resources and Environment, University of Electronic Science and Technology of China, Chengdu 611731, ChinaFull-waveform velocity inversion has long been a primary focus in seismic exploration. Full-waveform inversion techniques employing physics-informed recurrent neural networks (PIRNNs) have recently gained significant scholarly attention. However, these approaches demand considerable storage to capture spatiotemporal seismic wave propagation fields and their gradient information, often exceeding the memory capabilities of current GPU resources during field data processing. This study proposes a full-waveform inversion method utilizing a dual-branch PIRNN architecture to effectively minimize GPU resource consumption. The primary PIRNN branch performs forward-wave equation modeling at the original scale and computes the discrepancy between synthetic and observed seismic records. Additionally, a downscaled spatiotemporal PIRNN branch is introduced, transforming the original-scale error into a loss function via scale decomposition, which drives the inversion process in the downscaled domain. This dual-branch design necessitates recording only the spatiotemporal field and gradient information of the downscaled branch, significantly reducing GPU memory requirements. The proposed dual-branch PIRNN framework was validated through full-waveform inversions on synthetic horizontal-layer models and the Marmousi model across various scales. The results demonstrate that this approach markedly reduces resource consumption while maintaining high inversion accuracy.https://www.mdpi.com/2076-3417/15/2/941AFWImultiscale decompositiondual-physics-informed recurrent neural network (PIRNN)full-waveform inversion (FWI)vertical seismic profiling (VSP) |
| spellingShingle | Cai Lu Jijun Liu Liyuan Qu Jianbo Gao Hanpeng Cai Jiandong Liang Resource-Efficient Acoustic Full-Waveform Inversion via Dual-Branch Physics-Informed RNN with Scale Decomposition Applied Sciences AFWI multiscale decomposition dual-physics-informed recurrent neural network (PIRNN) full-waveform inversion (FWI) vertical seismic profiling (VSP) |
| title | Resource-Efficient Acoustic Full-Waveform Inversion via Dual-Branch Physics-Informed RNN with Scale Decomposition |
| title_full | Resource-Efficient Acoustic Full-Waveform Inversion via Dual-Branch Physics-Informed RNN with Scale Decomposition |
| title_fullStr | Resource-Efficient Acoustic Full-Waveform Inversion via Dual-Branch Physics-Informed RNN with Scale Decomposition |
| title_full_unstemmed | Resource-Efficient Acoustic Full-Waveform Inversion via Dual-Branch Physics-Informed RNN with Scale Decomposition |
| title_short | Resource-Efficient Acoustic Full-Waveform Inversion via Dual-Branch Physics-Informed RNN with Scale Decomposition |
| title_sort | resource efficient acoustic full waveform inversion via dual branch physics informed rnn with scale decomposition |
| topic | AFWI multiscale decomposition dual-physics-informed recurrent neural network (PIRNN) full-waveform inversion (FWI) vertical seismic profiling (VSP) |
| url | https://www.mdpi.com/2076-3417/15/2/941 |
| work_keys_str_mv | AT cailu resourceefficientacousticfullwaveforminversionviadualbranchphysicsinformedrnnwithscaledecomposition AT jijunliu resourceefficientacousticfullwaveforminversionviadualbranchphysicsinformedrnnwithscaledecomposition AT liyuanqu resourceefficientacousticfullwaveforminversionviadualbranchphysicsinformedrnnwithscaledecomposition AT jianbogao resourceefficientacousticfullwaveforminversionviadualbranchphysicsinformedrnnwithscaledecomposition AT hanpengcai resourceefficientacousticfullwaveforminversionviadualbranchphysicsinformedrnnwithscaledecomposition AT jiandongliang resourceefficientacousticfullwaveforminversionviadualbranchphysicsinformedrnnwithscaledecomposition |