The, role of feed spacers in membrane technology: 45 years of research

Separation and Purification Technology, Journal Year: 2024, Volume and Issue: 357, P. 130109 - 130109

Published: Oct. 19, 2024

Language: Английский

---

Prediction of Coalbed Methane Production Using a Modified Machine Learning Methodology

Energies, Journal Year: 2025, Volume and Issue: 18(6), P. 1341 - 1341

Published: March 9, 2025

Compared to natural and shale gas, studies on predicting production specific coalbed methane (CBM) are still relatively limited, mainly use decline curve methods such as Arps, Stretched Exponential Decline Model, Duong’s model. In recent years, machine learning (ML) applied CBM prediction have focused the significant data characteristics of production, achieving more accurate predictions. However, throughout application process, these models require a large amount for training can only achieve forecasts over short period, 30 days. This study constructs hybrid ML model by integrating long short-term memory (LSTM) network Transformer architecture. The is trained using mean absolute error (MAE) loss function, optimized Adam optimizer, finally evaluated metrics MAE, root square (RMSE), R squared (R2) scores. results show that LSTM-Attention (LSTM-A) based small datasets accurately capture trend superior traditional LSTM regarding accuracy effective time interval. methodologies established obtained in this great significance predict production. It also helpful better understand mechanisms

Language: Английский

---

Effect of tin filler composition on porosity in tin-polydimethylsiloxane composites

Pure and Applied Chemistry, Journal Year: 2025, Volume and Issue: unknown

Published: April 8, 2025

Abstract A high porosity in radiation shielding material led to penetrating, raising the exposure risk for workers, patients, and public. Thus, this study is designed observe evaluate morphology structure of a composite its porosity. Tin-PDMS-based prepared by dispersing pure tin powder into PDMS polymer liquid at different weight percentages powder, 10 %, 20 30 40 50 60 %. It was analysed under Field Emission Scanning Electron Microscopy (FESEM), energy dispersive X-ray (EDX) evaluated with ImageJ software. FESEM showed an intact low porosity, Fourier Transform Infrared Spectroscopy (FTIR) analysis verified that had been successfully incorporated matrix. The material’s compositional integrity confirmed EDX analysis, which revealed progressive increase content along decrease oxygen silicon concentrations. With % filler showing maximum 0.34 measurements small rise increasing compositions. exhibited highest pore size (0.031 µm), indicating doesn’t higher metal content. Therefore, novelty lies optimisation dispersion within achieve effective balance between attenuation capability ensure compact can attenuate beam successfully.

Language: Английский

---

Pore-scale investigation of bubble evolution in shallow coastal gassy silt with micro-computed tomography

Applied Ocean Research, Journal Year: 2025, Volume and Issue: 159, P. 104609 - 104609

Published: May 17, 2025

Language: Английский

---

Analysis of Microscopic Remaining Oil Based on the Fluorescence Image and Deep Learning

Journal of Fluorescence, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 18, 2024

Language: Английский

---

Carbonate reservoirs characterization based on frequency Bayesian principal component analysis

Geoenergy Science and Engineering, Journal Year: 2024, Volume and Issue: unknown, P. 213615 - 213615

Published: Dec. 1, 2024

Language: Английский

---

Lightweight multi-scale attention feature distillation network for super-resolution reconstruction of digital rock images

Geoenergy Science and Engineering, Journal Year: 2024, Volume and Issue: unknown, P. 213628 - 213628

Published: Dec. 1, 2024

Language: Английский

---