Applicability of Anomaly Detection Knowledge from Computer Science and Mathematics Publications to Oil and Gas Research

Abstract

Anomaly detection in equipment processes is crucial for the oil and gas sector. Algorithms for detecting anomalies in measured data are best understood in Computer Science and Mathematics. Therefore, a possible transfer of knowledge from the latter area to the former can have a profound impact. This paper explores the potential for the knowledge transfer by analyzing bibliometric data of Computer Science and Mathematics papers published in MDPI journals, as well as publications available on SPE search platform. The research shows that the main algorithms extensively studied in MDPI publications and found in SPE publications, which address the anomaly detection problem, include Random Forest, Support Vector Machine, Long-term Memory Method and Recurrent Neural Network. The main advantages and disadvantages of these methods are briefly described. This paper provides examples of classical and highly cited publications that describe the work of these algorithms along with examples of articles that illustrate their application in the oil and gas industry. The sections of SPE disciplines with the largest number of publications using the algorithms that are frequently used for anomaly detection are presented.