CONTACT US: Contact info
|
Citation |
Park JW, Kim W, Jung BC, Moon SJ, Yang SJ. Transactions of the Korean Society of Mechanical Engineers, A 2021; 45(11): 1019-1028. |
|
Copyright |
(Copyright © 2021) |
|
DOI |
|
PMID |
unavailable |
|
Abstract |
In various industrial sites, pipeline damages due to external shock, pipeline corrosion, mechanical damage, and consequent severe accidents have been reported. To quickly detect and isolate a damaged pipeline when rupture occurs, we research smart valves with a support vector machine (SVM)-based damage detection algorithm. First, a testbed to validate the developed damage detection algorithm is built, and pressure data in various damage conditions are collected. Second, the main features for damage detection are selected and classifiers are designed based on the SVM. Finally, the designed classifier is installed on the smart valves. The method's effectiveness for autonomous detection and isolation of the damage in the testbed is confirmed. A smart valve with SVM-based damage detection algorithm is expected be helpful in reducing maintenance cost in pipeline systems. © 2021 Korean Society of Mechanical Engineers. Language: en |
|
Keywords |
Pipeline corrosion; Pipelines; Support vector machines; Damage detection; Signal detection; Testbeds |