An overlapping peak separation algorithm based on multiorder differential method and genetic algorithm for magnetic eddy current signal of a defect cluster. (31st December 2019)

• Record Type: Journal Article
• Authors: Xiong, Jingyi; Liang, Wei; Liang, Xiaobin; Zhang, Meng
• Other Names: Nasif Mohammad Shakir guestEditor.; Rusli Risza Bt guestEditor.; Willey Ron guestEditor.
• Is Part Of: Process safety progress. Volume 39(2020)Supplement 1
• Page Start: n/a
• Page End: n/a
• Publication Date: 2019-12-31
• Subjects: data fitting -- defect cluster -- genetic algorithm -- magnetic eddy current testing -- multiorder differential; Chemical plants -- Management -- Periodicals; 660
• DOI: 10.1002/prs.12129
• Languages: English
• ISSNs: 1066-8527

Abstract

Abstract: Safety assessment plays a vital role in the operation of oil and gas production systems, and data processing is important for the analysis of defects to make correct diagnosis. Therefore, the research to find a reliable data processing method is carried out. The magnetic eddy current signal detected of the defect is generally an abnormal data of "one peak and double valley" type. When the distance between the two defects is close, the leakage magnetic fields of the two defects interfere with each other. In order to facilitate the extraction of the characteristics of such anomalous data, it is necessary to separate the overlapping abnormal data. In this article, the above methods for identifying and processing anomaly detection data of complex defects are studied. The data fitting method is used to find the most suitable fitting function, and the accuracy of overlapping peak separation is optimized by the improved peak separation method based on GA algorithm. The results show that the Gaussian function is most suitable for fitting prediction of the "defective cluster" detection data after separation. The overlapped peak separation result optimized by GA algorithm has less error with the actual data. Therefore, the relevant features of the separated data can accurately reflect the defect related information and effectively improve the pipeline safety assessment.