Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
10321946 | Expert Systems with Applications | 2013 | 10 Pages |
Abstract
⺠This paper presents an oil and gas pipeline failure prediction system using LRUT in conjunction with Euclidean-SVM approach. ⺠The conventional NDT pipeline failure prediction systems are deployed at pre-determined intervals. ⺠LRUT pipeline failure prediction system with Euclidean-SVM provides continuous monitoring for pipelines and makes decision without human errors and misinterpretations. ⺠In contrast to the conventional SVM, Euclidean-SVM is less dependent on the choice of the kernel function and parameters. ⺠Experimental results show that Euclidean-SVM has more consistent and better classification performance as compared to the conventional SVM.
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Lam Hong Lee, Rajprasad Rajkumar, Lai Hung Lo, Chin Heng Wan, Dino Isa,