Title: Defect feature extraction and recognition of buried pipeline based on metal magnetic memory

Authors: Yong Yang; Guan-Jun Wang; Yu Wang; Yong Wan; Yong-Shou Dai

Addresses: Technology Inspection Center, ShengLi Oil Field, ShanDong Dongying, 257000, China ' Technology Inspection Center, ShengLi Oil Field, ShanDong Dongying, 257000, China ' College of Information and Control Engineering, China University of Petroleum, Qingdao, 266580, China ' College of Information and Control Engineering, China University of Petroleum, Qingdao, 266580, China ' College of Information and Control Engineering, China University of Petroleum, Qingdao, 266580, China

Abstract: The surfaces of metal pipelines are always susceptible to various types of defects and damages, including corrosion defects and early stress concentration defects. Metal magnetic memory detection technology is the only non-destructive testing technology that can diagnose the early damage of ferromagnetic components. However, the metal magnetic memory original signal itself cannot directly recognise and distinguish corrosion defects and stress concentration defects. To solve this problem, this paper establishes a multi-characteristic statistical recognition method for the two defect types based on the metal magnetic memory technology and the magnetic memory test data obtained from pipeline test pieces. Next, this method is used to identify the defect types of four pipelines in the oil field environment; the results demonstrate that the established defect type recognition method is effective for the identification of pipeline corrosion defects and early stress concentration defects.

Keywords: metal magnetic memory; pipeline defects; corrosion defects; stress concentration defects; feature extraction; defect recognition; signal processing.

DOI: 10.1504/IJMIC.2020.114789

International Journal of Modelling, Identification and Control, 2020 Vol.35 No.4, pp.353 - 362

Received: 06 May 2020
Accepted: 01 Jul 2020

Published online: 29 Apr 2021 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article