The research progress and prospect of data mining methods on corrosion prediction of oil and gas pipelines. (February 2023)

• Record Type: Journal Article
• Authors: Xu, Lei; Wang, Yunfu; Mo, Lin; Tang, Yongfan; Wang, Feng; Li, Changjun
• Is Part Of: Engineering failure analysis. Volume 144(2023)
• Publication Date: 2023-02
• Subjects: Machine learning -- Decision support -- Oil and gas pipeline systems -- Corrosion data -- Corrosion control effect; System failures (Engineering) -- Periodicals; Fracture mechanics -- Periodicals; Reliability (Engineering) -- Periodicals; Pannes -- Périodiques; Rupture, Mécanique de la -- Périodiques; Fiabilité -- Périodiques; Fracture mechanics; Reliability (Engineering); System failures (Engineering); Periodicals; Electronic journals; 620.112
• DOI: 10.1016/j.engfailanal.2022.106951
• Languages: English
• ISSNs: 1350-6307

Abstract

Highlights: By constructing the application framework of corrosion prediction, this paper indicates the necessity of data preprocessing and feature analysis. Random forest and deep learning have extensive application prospects in this field. Machine learning can effectively achieve intelligent corrosion prediction and improve corrosion control effect. Abstract: As the principal means of oil and natural gas transportation, oil and gas pipeline systems suffer from common corrosion problems, accurate corrosion prediction of oil and gas pipelines has an essential influence on pipe material selection, remaining useful life prediction, maintenance planning, etc. At present, a large number of corrosion monitoring techniques are applied in oil and gas pipeline systems. The monitored data have the characteristics of multidimensional quantities, noise interference, non-linearity, etc. Machine learning can effectively solve the limitations of relying solely on mathematical models to achieve intelligent corrosion prediction and improve the corrosion control effect. Considering the corrosion prediction problems in oil and gas pipeline systems, the application of machine learning methods in corrosion rate prediction, oil and gas pipeline leakage and defect assessment, and corrosion image recognition were focused on in this paper. By constructing the application framework of machine learning in the field of oil and gas pipeline corrosion prediction, the necessity of data preprocessing and feature correlation analysis are indicated in this paper. Furthermore, random forest and deep learning have extensive application prospects in this field. Finally, the application prospects of machine learning were discussed.