آمار

تعداد نشریات21
تعداد شماره‌ها356
تعداد مقالات3,977
تعداد مشاهده مقاله5,605,139
تعداد دریافت فایل اصل مقاله3,749,819
Farasat, Mojtaba, Sori, Abobakr, Rahimi, Mojtaba, Farasat, Amir, Eftekhari, Seyyed Ali. (1404). Improving the accuracy of critical downhole equipment detection using unsupervised machine learning and deep learning methods. , 4(1), 38-52. doi: 10.22077/jgm.2026.11458.1080
Mojtaba Farasat; Abobakr Sori; Mojtaba Rahimi; Amir Farasat; Seyyed Ali Eftekhari. "Improving the accuracy of critical downhole equipment detection using unsupervised machine learning and deep learning methods". , 4, 1, 1404, 38-52. doi: 10.22077/jgm.2026.11458.1080
Farasat, Mojtaba, Sori, Abobakr, Rahimi, Mojtaba, Farasat, Amir, Eftekhari, Seyyed Ali. (1404). 'Improving the accuracy of critical downhole equipment detection using unsupervised machine learning and deep learning methods', , 4(1), pp. 38-52. doi: 10.22077/jgm.2026.11458.1080
Farasat, Mojtaba, Sori, Abobakr, Rahimi, Mojtaba, Farasat, Amir, Eftekhari, Seyyed Ali. Improving the accuracy of critical downhole equipment detection using unsupervised machine learning and deep learning methods. , 1404; 4(1): 38-52. doi: 10.22077/jgm.2026.11458.1080

Improving the accuracy of critical downhole equipment detection using unsupervised machine learning and deep learning methods

Journal of Geomine
دوره 4، شماره 1، خرداد 2026، صفحه 38-52    اصل مقاله (1.16 M)
نوع مقاله: Original Article
شناسه دیجیتال (DOI): 10.22077/jgm.2026.11458.1080
نویسندگان
Mojtaba Farasat1؛ Abobakr Sori2؛ Mojtaba Rahimi* 3؛ Amir Farasat4؛ Seyyed Ali Eftekhari5
1Department of Petroleum Engineering, Kho.C., Islamic Azad University, Khomeinishahr, Iran
2Transport Phenomena Research Center, Chemical Engineering Faculty, Sahand University of Technology, P.O. Box 51335/1996 Tabriz, Iran
3Department of Petroleum Engineering, Kho.C., Islamic Azad University, Khomeinishahr, Iran; Stone Research Center, Kho.C., Islamic Azad University, Khomeinishahr, Iran
4Iranian Offshore Oil Company, Tehran, Iran
5Department of Mechanical Engineering, Kho.C., Islamic Azad University, Khomeinishahr, Iran
چکیده
The subsurface equipment industry is crucial to sustainable oil extraction, ensuring efficiency, safety, and the absence of unplanned shutdowns. Conventional approaches to determining critical equipment are becoming less effective in the face of the current amount and complexity of operational data. This study presents a new unsupervised technique using deep clustering to fill this research gap. Based on actual sensor data from Electric Submersible Pumps (ESPs) over two years, the approach entails extensive preprocessing, such as cleaning, normalization, and handling missing values. The proposed system uses a deep clustering method based on a 1D convolutional autoencoder neural network and validates its performance against the traditional K-Means algorithm. The quality of the clusters was assessed using the objective function (OBJ) and mean squared distance (MSD) measures. The outcome proves the proposed approach to be highly superior, with an MSD of 845 and an OBJ of 6.23, compared to 1386 and 0.96 for K-Means. This shows a highly improved cluster density and data point cohesion. Further analysis revealed specific equipment behavior patterns, where Cluster 4 corresponds to critical and unstable equipment that needs to be addressed immediately, and Cluster 3 corresponds to stable and low-risk equipment. This paper proves that deep clustering with the help of autoencoders is a precise and efficient technique for critical equipment detection in a complex industrial setting, with immense potential for optimizing predictive maintenance approaches in the oil and gas industry.
کلیدواژه‌ها
Deep Clustering؛ Predictive Maintenance؛ Critical Equipment Detection؛ Unsupervised Learning؛ Oil and Gas Industry
آمار
تعداد مشاهده مقاله: 34
تعداد دریافت فایل اصل مقاله: 17


نماد الکترونیک 

 

سامانه مدیریت نشریات علمی. قدرت گرفته از سیناوب