A Novel Rate of Penetration Fusion Modeling Method based on Improved Dung Beetle Optimization Algorithm and Support Vector Regression

The rate of penetration (ROP) is a critical indi-cator for evaluating drilling efficiency. Developing an accurate ROP model is essential for optimizing drilling performance and addressing process control challenges. However, ROP modeling in deep geological drilling is complicated by nonlinearity, di...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE International Conference on Industrial Technology (Online) S. 1 - 6
Hauptverfasser: Zhou, Yang, Zhang, Zihan, Chen, Shipeng, Wu, Min, Chen, Xin
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 26.03.2025
Schlagworte:
Benchmark testing
Chaos
Clustering algorithms
Clustering methods
Data models
deep geo-logical drilling process
Drilling
Employee welfare
fuzzy C-means
Geology
improved dung beetle optimization
Optimization
Rate of penetration
Support vector machines
support vector regression
ISSN:2643-2978
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The rate of penetration (ROP) is a critical indi-cator for evaluating drilling efficiency. Developing an accurate ROP model is essential for optimizing drilling performance and addressing process control challenges. However, ROP modeling in deep geological drilling is complicated by nonlinearity, diverse working conditions, and high-dimensional variations. To overcome these challenges, a fusion modeling approach for ROP is proposed. First, the fuzzy C-means clustering method is applied to classify drilling data into different working con-ditions. Based on this classification, support vector regression is employed to develop ROP sub-models, effectively addressing nonlinearity. To further enhance model accuracy, an improved dung beetle optimization algorithm (IDBO) is designed to deter-mine optimal model parameters and integrate the sub-models, thereby resolving issues related to multiple working conditions and high-dimensional variations. The IDBO incorporates four key enhancements, average weight, chaos disturbance, modified local search, and re-updating of the best solution, to strength-en its global search capability. Comparative results using the IEEE CEC2017 benchmark test functions demonstrate that the proposed algorithm outperforms others in 12 test functions, highlighting its strong global optimization ability. Additionally, results from real-world drilling data validate the effectiveness of the proposed modeling approach in practical applications.
ISSN:2643-2978
DOI:10.1109/ICIT63637.2025.10965266