Detalhes bibliográficos
| Ano de defesa: |
2023 |
| Autor(a) principal: |
DIAS JÚNIOR, Domingos Alves
 |
| Orientador(a): |
SILVA, Aristófanes Corrêa
|
| Banca de defesa: |
SILVA, Aristófanes Corrêa
,
PAIVA, Anselmo Cardoso de
,
ALMEIDA, João Dallyson Sousa de
,
CUNHA, António Manuel Triguieros da Silva |
| Tipo de documento: |
Dissertação
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Universidade Federal do Maranhão
|
| Programa de Pós-Graduação: |
PROGRAMA DE PÓS-GRADUAÇÃO EM CIÊNCIA DA COMPUTAÇÃO/CCET
|
| Departamento: |
DEPARTAMENTO DE ENGENHARIA DA ELETRICIDADE/CCET
|
| País: |
Brasil
|
| Palavras-chave em Português: |
|
| Palavras-chave em Inglês: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
https://tedebc.ufma.br/jspui/handle/tede/4574
|
Resumo: |
Seismic reflection is one of the geophysical methods most used in the oil and gas (O&G) industry for hydrocarbon prospecting. In particular, for some Brazilian onshore fields, such a method has been used for estimating the location and volume of gas accumulations. However, the analysis and interpretation of seismic data are time-consuming due to the large amount of information and the noisy nature of the acquisitions. Then, to support geoscientists in those tasks, computational tools powered by machine learning have been proposed to detect potential gas accumulations. In this study, we proposed a method organized into two stages: (1) pre-processing applied to the image database (delimitation of the region of interest, clustering regions of seismic images, generation of space-time samples, and sample optimization); and (2) detection of gas accumulations based on the Convolutional Long Short-Term Memory (ConvLSTM) model. Experiments were performed on seismic reflection images from exploration fields belonging to the Paranaíba basin. Then, the best scenarios achieved an F1-score of 58.11%, a sensitivity of 83.36%, a precision 44.63% , a specificity of 98.43% and an accuracy of 99.29% in the Preto exploration field . Then, the Real field achieved an F1-score of 60.4%, a sensitivity of 77.79% , a precision of 49,36%, a specificity of 98,62% and an accuracy of 99,38%. Besides, the Branco field achieved an F1-score 60.14% , a sensitivity of 77.89% , a precision of 48.98%, specificity of 96,93% and an accuracy of 99,66%. Finally, 85.51%, 98.88%, 75.33%, 99.30% and 99.35% of F1-score, sensitivity, precision, specificity and accuracy were obtained for the Vermelho field, respectively. In summary, the results provide strong evidence that the proposed method is a tool with potential to detect potential gas accumulations. |
