EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Drilling Parameters Multi-Objective Optimization Method Based on PSO-Bi-LSTM

Jianhua Wang, Zhi Yan, Tao Pan, Zhaopeng Zhu, Xianzhi Song, Donghan Yang

Applied Sciences · 25.10.2023 · 13 (21) · 11666

DOI: 10.3390/app132111666 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Applied Sciences, online erschienen am 25.10.2023. Verfasst von Jianhua Wang und 5 weiteren Autor:innen. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Strategic Cooperation Technology Projects of CNPC and CUPB, National Key Research and Development Program, National Science Fund for Distinguished Young Scholars.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The increasing exploration and development of complex oil and gas fields pose challenges to drilling efficiency and safety due to the presence of formations with varying hardness, abrasiveness, and rigidity. Consequently, there is a growing demand for drilling parameter optimization and speed-up technologies. However, existing models based on expert experience can only achieve single-objective optimization with limited accuracy, making real-time adaptation to changing drilling conditions and formation environments challenging. The emergence of artificial intelligence provides a new approach for optimizing drilling parameters. In this study, we introduce the Bi-directional Long Short-Term Memory (Bi-LSTM) deep learning algorithm with the attention mechanism to predict the rate of penetration (ROP). This algorithm improves the ROP prediction accuracy to 98.33%, ensuring reliable subsequent optimization results. Additionally, we propose a coupling optimization algorithm that combines Bi-LSTM with the particle swarm optimization algorithm (PSO) to enhance drilling efficiency through parameter optimization. Our approach aims to maximize drilling footage while maintaining the highest ROP. The optimal solutions obtained are verified through multi-parameter cloud image analysis, yielding consistent results. The application of our approach demonstrates an 81% increase in drilling speed and a 28% reduction in drill bit energy losses. Moreover, the real-time optimization results effectively guide field operations.

Förderung

  • Strategic Cooperation Technology Projects of CNPC and CUPB
  • National Key Research and Development Program
  • National Science Fund for Distinguished Young Scholars
  • China Petroleum Innovation Fund Project
  • Science Foundation of China University of Petroleum, Beijing

Zitieren

APA

Wang, J., Yan, Z., Pan, T., Zhu, Z., Song, X., & Yang, D. (2023). Drilling Parameters Multi-Objective Optimization Method Based on PSO-Bi-LSTM. Applied Sciences, 13(21), 11666. https://doi.org/10.3390/app132111666

Vancouver

Wang J, Yan Z, Pan T, Zhu Z, Song X, Yang D. Drilling Parameters Multi-Objective Optimization Method Based on PSO-Bi-LSTM. Applied Sciences. 2023;13(21):11666. doi:10.3390/app132111666

BibTeX
@article{Wang2023,
  author = {Wang, Jianhua and Yan, Zhi and Pan, Tao and Zhu, Zhaopeng and Song, Xianzhi and Yang, Donghan},
  title = {Drilling Parameters Multi-Objective Optimization Method Based on PSO-Bi-LSTM},
  journal = {Applied Sciences},
  year = {2023},
  volume = {13},
  number = {21},
  pages = {11666},
  doi = {10.3390/app132111666}
}
Kostenlos starten →