Active incremental Support Vector Machine for oil and gas pipeline defects prediction system using long range ultrasonic transducers

• Published in: Ultrasonics Vol. 54; no. 6; pp. 1534 - 1544
• Main Authors: Akram, Nik Ahmad, Isa, Dino, Rajkumar, Rajprasad, Lee, Lam Hong
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.08.2014
• Subjects: Artificial intelligence; Classification; Defects; Gas pipelines; Incremental learning; Non-Destructive Testing; Oil and gas pipeline; Petroleum pipelines; Pigs; Sensors; Support Vector Machine; Support vector machines; Transducers; Incremental learning; Oil and gas pipeline; Support Vector Machine; Artificial intelligence; Non-Destructive Testing
• ISSN: 0041-624X; 1874-9968
• DOI: 10.1016/j.ultras.2014.03.017

•LRUT and active incremental SVM are proposed for a real-time pipeline defect prediction system.•The model is evaluated using data acquired from lab scale pipeline rig using LRUT.•Results show that active-set method reduces the execution time for incremental trained SVM.•The proposed method demonstrates shorter execution time as compared to the selected incremental SVM methods.•Results in term of accuracy validate that incremental SVM is applicable for LRUT based NDT system. This work proposes a long range ultrasonic transducers technique in conjunction with an active incremental Support Vector Machine (SVM) classification approach that is used for real-time pipeline defects prediction and condition monitoring. Oil and gas pipeline defects are detected using various techniques. One of the most prevalent techniques is the use of “smart pigs” to travel along the pipeline and detect defects using various types of sensors such as magnetic sensors and eddy-current sensors. A critical short coming of “smart pigs” is the inability to monitor continuously and predict the onset of defects. The emergence of permanently installed long range ultrasonics transducers systems enable continuous monitoring to be achieved. The needs for and the challenges of the proposed technique are presented. The experimental results show that the proposed technique achieves comparable classification accuracy as when batch training is used, while the computational time is decreased, using 56 feature data points acquired from a lab-scale pipeline defect generating experimental rig.