A Sequence Learning Approach for Real-Time and Ahead-of-Bit Pore Pressure Prediction Utilizing Drilling Data from the Drilled Section DOI
Heng Yang, Yongcun Feng,

Guanyi Shang

et al.

SPE Journal, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 20

Published: Dec. 1, 2024

Summary Accurate pore pressure prediction is vital for ensuring drilling safety and efficiency. Existing methods primarily rely on interpreting logging while (LWD) data real-time prediction. However, LWD tools typically collect from sensors located approximately 100 ft behind the drill bit, reflecting formations that have already been penetrated rather than those being actively drilled. In contrast, reflect drilled at without requiring additional downhole equipment or extra costs. Nevertheless, traditional using often employ simplified theoretical formulas oversimplify complex characteristics of geological conditions. Although a few studies utilized machine learning with prediction, they point-to-point methods, given depth to predict same depth. This approach overlooks sequential nature along well depth, limiting accuracy ability forecast ahead which crucial proactive decision-making. Therefore, this study proposed novel utilizes historical upper section (drilled window) pressure, specifically employing two methods: (1) Real-time predictions use sequence-to-point strategy, where window are used bit. (2) Ahead-of-bit sequence-to-sequence undrilled developed three custom-designed neural network models long short-term memory (LSTM) self-attention algorithms: LSTM, Double-Layer LSTM-Attention. For LSTM model 15-m length achieves stable performance mean squared error (MSE) 1.45×10⁻⁴. Integrating bit further improves accuracy, increasing coefficient determination (R²) 0.61 0.89 Well Test-1 0.50 0.75 Test-2. Field tests ongoing wells demonstrate practicality robustness approach, achieving R² values 0.72 0.83. ahead-of-bit provides reference guidance distances 10, 20, 30, 40 m presenting optimal configurations each scenario. The LSTM-Attention demonstrates superior performance. as distance increases, also grows. recommended configuration set 30 80 m, yielding an MSE 2.88×10⁻⁴. strikes balance between distance, maximum maintaining acceptable level accuracy. operators can flexibly choose based their specific requirements distance. could achieve accurate predictions, facilitating early identification risks enabling timely adjustments, thereby improving

Language: Английский

Advanced thermal prediction for green roofs: CNN-LSTM model with SSA optimization DOI
Jun Wang, Xu Ding,

Wansheng Yang

et al.

Energy and Buildings, Journal Year: 2024, Volume and Issue: 322, P. 114745 - 114745

Published: Aug. 31, 2024

Language: Английский

Citations

9
Using the TSA-LSTM two-stage model to predict cancer incidence and mortality DOI Creative Commons
Rabnawaz Khan, Jie Wang

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(2), P. e0317148 - e0317148

Published: Feb. 20, 2025

Cancer, the second-leading cause of mortality, kills 16% people worldwide. Unhealthy lifestyles, smoking, alcohol abuse, obesity, and a lack exercise have been linked to cancer incidence mortality. However, it is hard. Cancer lifestyle correlation analysis mortality prediction in next several years are used guide people's healthy lives target medical financial resources. Two key research areas this paper Data preprocessing sample expansion design Using experimental comparison, study chooses best cubic spline interpolation technology on original data from 32 entry points 420 converts annual into monthly solve problem insufficient prediction. Factor possible because sources indicate changing factors. TSA-LSTM Two-stage attention popular tool with advanced visualization functions, Tableau, simplifies paper's study. Tableau's testing findings cannot analyze predict time series data. LSTM utilized by optimization model. By commencing input feature attention, model technique guarantees that encoder converges subset sequence features during output features. As result, model's natural learning trend quality enhanced. The second step, performance maintains We can choose network improve forecasts based real-time performance. Validating source factor using Most cancers overlapping risk factors, excessive drinking, exercise, obesity breast, colorectal, colon cancer. A poor directly promotes lung, laryngeal, oral cancers, according visual tests. expected climb 18-21% between 2020 2025, 2021. Long-term projection accuracy 98.96 percent, smoking may be main causes.

Language: Английский

Citations

0
iCNN-LSTM: An Incremental CNN-LSTM Based Ransomware Detection System DOI
Jamil Ispahany, Rafiqul Islam, Md Zahidul Islam

et al.

Lecture notes in computer science, Journal Year: 2025, Volume and Issue: unknown, P. 46 - 60

Published: Jan. 1, 2025

Language: Английский

Citations

0
PSO-Optimized Data-Driven and Mechanism Hybrid Model to Enhance Prediction of Industrial Hydrocracking Product Yields Under Data Constraints DOI Open Access
Zhenming Li, Kang Qin, Yang Zhang

et al.

Processes, Journal Year: 2025, Volume and Issue: 13(4), P. 1118 - 1118

Published: April 8, 2025

The accurate prediction of hydrocracking product yields is crucial for optimizing resource allocation and improving production efficiency. However, the flowrates in units often faces challenges due to insufficient data weak correlations between input output variables. This study proposes a hybrid framework combining Convolutional Neural Network–Long Short-Term Memory (CNN-LSTM) model, mechanism modeling, Particle Swarm Optimization (PSO) address these issues. CNN-LSTM captures spatiotemporal dependencies operational data, while model incorporates domain-specific physical constraints. structured both series parallel configurations, with PSO key hyperparameters enhance its predictive performance. results demonstrate significant improvements accuracy, determination coefficients (R2s) reaching 0.896 (kerosene), 0.879 (residue), 0.899 (heavy naphtha), 0.78 (light naphtha). Shapley Additive Explanations (SHAP) Mutual Information Coefficient (MIC) analyses highlight model’s role feature interpretability. underscores efficacy integrating kinetics deep learning, metaheuristic optimization complex industrial processes under constraints, offering robust approach yield prediction.

Language: Английский

Citations

0
Data-driven dynamic inclination angle estimation of monorail crane under complex road conditions DOI
Zechao Liu, Weimin Wu, Jingzhao Li

et al.

Measurement Science and Technology, Journal Year: 2024, Volume and Issue: 35(11), P. 116117 - 116117

Published: Aug. 12, 2024

Abstract Monorail cranes are crucial in facilitating auxiliary transportation within deep mining operations. As unmanned driving technology becomes increasingly prevalent monorail crane operations, it encounters challenges such as low accuracy and unreliable attitude recognition, significantly jeopardizing the safety of Hence, this study proposes a dynamic inclination estimation methodology utilizing Estimation-Focused-EKFNet algorithm. Firstly, based on characteristics crane, model is established, which value can be calculated real-time by extended Kalman filter (EKF) estimator; however, given complexity road conditions, order to improve recognition accuracy, CNN-LSTM-ATT algorithm combining convolutional neural network (CNN), long short-term memory (LSTM) attention mechanism (ATT) used firstly predict current camber predicted combined with CNN mechanism, then observation EKF estimator, finally realizes that estimator output accurate real-time. Experimental results indicate that, compared unscented filter, LSTM-ATT, CNN-LSTM algorithms, enhances complex conditions at least 52.34%, improving reliability. Its reaches 99.28%, effectively ensuring for cranes.

Language: Английский

Citations

3
Harnessing Deep Learning and Snow Cover Data for Enhanced Runoff Prediction in Snow-Dominated Watersheds DOI Creative Commons
Rana Muhammad Adnan Ikram, Mo Wang, Özgür Kişi

et al.

Atmosphere, Journal Year: 2024, Volume and Issue: 15(12), P. 1407 - 1407

Published: Nov. 22, 2024

Predicting streamflow is essential for managing water resources, especially in basins and watersheds where snowmelt plays a major role river discharge. This study evaluates the advanced deep learning models accurate monthly peak forecasting Gilgit River Basin. The utilized were LSTM, BiLSTM, GRU, CNN, their hybrid combinations (CNN-LSTM, CNN-BiLSTM, CNN-GRU, CNN-BiGRU). Our research measured model’s accuracy through root mean square error (RMSE), absolute (MAE), Nash–Sutcliffe efficiency (NSE), coefficient of determination (R2). findings indicated that models, CNN-BiGRU achieved much better performance than traditional like LSTM GRU. For instance, lowest RMSE (71.6 training 95.7 testing) highest R2 (0.962 0.929 testing). A novel aspect this was integration MODIS-derived snow-covered area (SCA) data, which enhanced model substantially. When SCA data included, CNN-BiLSTM improved from 83.6 to 71.6 during 108.6 testing. In prediction, outperformed other with (108.4), followed by (144.1). study’s results reinforce notion combining CNN’s spatial feature extraction capabilities temporal dependencies captured or GRU significantly enhances accuracy. demonstrated improvements prediction accuracy, extreme events, highlight potential these support more informed decision-making flood risk management allocation.

Language: Английский

Citations

2
Projected waste and recycling potential of China’s photovoltaic industry DOI
Bingchun Liu, Ming Li, Jiali Chen

et al.

Waste Management, Journal Year: 2024, Volume and Issue: 191, P. 264 - 273

Published: Nov. 20, 2024

Language: Английский

Citations

1
Automatic assessment of CFRP-steel interfacial performance under adhesive curing using PZT-based EMI-integrated deep learning technique DOI
Jun Deng, X.J. Wu,

Xiaoda Li

et al.

Thin-Walled Structures, Journal Year: 2024, Volume and Issue: unknown, P. 112894 - 112894

Published: Dec. 1, 2024

Language: Английский

Citations

1
Identification and Prediction of Casing Collar Signal Based on CNN-LSTM DOI
Jun Jing,

Yiman Qin,

Xiaohua Zhu

et al.

Arabian Journal for Science and Engineering, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 16, 2024

Language: Английский

Citations

0
The State-of-the-Art Review on the Drill Pipe Vibration DOI
Jinze Song, Shuai Liu,

Yufa He

et al.

Geoenergy Science and Engineering, Journal Year: 2024, Volume and Issue: unknown, P. 213337 - 213337

Published: Sept. 1, 2024

Language: Английский

Citations

0