A comparative study of MLP and LSTM neural networks for shale gas production prediction based on numerical simulation data

Xiaoou Fei, Man Ye, Zhou Du and Haibin Miao

PLOS ONE, 2025, vol. 20, issue 11, 1-18

Abstract: Accurate prediction of shale gas production is essential for optimizing reservoir development and improving production efficiency. In this study, a numerical simulation model was first developed to systematically calculate daily shale gas production under various engineering parameter combinations, thereby establishing a comprehensive production prediction database. Two types of deep learning models—multi-layer perceptron (MLP) and long short-term memory (LSTM) neural networks—were then constructed to predict daily shale gas production. Comparisons with actual production data for three representative scenarios revealed that the MLP model achieved relative errors of 2.43%, 6.36%, and 4.16%, while the LSTM model achieved superior accuracy with relative errors of 0.42%, 1.1%, and 0.98%. The LSTM network’s gating mechanisms effectively captured the long-term dependencies in shale gas production data, making it more suitable for complex multi-scale dynamic modeling compared to the feedforward MLP. These results demonstrate the excellent generalization capability and engineering applicability of deep learning techniques, particularly LSTM networks, for enhancing shale gas production forecasting and supporting the efficient development of unconventional gas reservoirs.

Date: 2025
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0336782 (text/html)
https://journals.plos.org/plosone/article/file?id= ... 36782&type=printable (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:plo:pone00:0336782

DOI: 10.1371/journal.pone.0336782

Access Statistics for this article

More articles in PLOS ONE from Public Library of Science
Bibliographic data for series maintained by plosone ().