Data-driven Methods to Predict the Burst Strength of Corroded Line Pipelines Subjected to Internal Pressure

作     者：Jie Cai Xiaoli Jiang Yazhou Yang Gabriel Lodewijks Minchang Wang Jie Cai;Xiaoli Jiang;Yazhou Yang;Gabriel Lodewijks;Minchang Wang

作者机构：Department of Technology and InnovationUniversity of Southern DenmarkCampusvej 55-5230 OdenseDenmark Department of Maritime and Transport TechnologyDelft University of Technology2628 CD Delftthe Netherlands College of Advanced Interdisciplinary StudiesNational University of Defence TechnologyChangsha410073P.R.China School of EngineeringUniversity of NewcastleNSW2308Australia Hangzhou Silan Microelectronics Co.Ltd.HangzhouZhejiang310008P.R.China. 

出 版 物：《Journal of Marine Science and Application》 (船舶与海洋工程学报（英文版）)

年 卷 期：2022年第21卷第2期

页      面：115-132页

核心收录：

学科分类：08[工学] 0824[工学-船舶与海洋工程] 0823[工学-交通运输工程] 

主　　题：Pipelines Corrosion Burst strength Internal pressure Data-driven method Machine learning 

摘      要：A corrosion defect is recognized as one of the most severe phenomena for high-pressure pipelines,especially those served for a long ***-element method and empirical formulas are thereby used for the strength prediction of such pipes with ***,it is time-consuming for finite-element method and there is a limited application range by using empirical *** order to improve the prediction of strength,this paper investigates the burst pressure of line pipelines with a single corrosion defect subjected to internal pressure based on data-driven *** supervised ML(machine learning)algorithms,including the ANN(artificial neural network),the SVM(support vector machine)and the LR(linear regression),are deployed to train models based on experimental *** analysis is first conducted to determine proper pipe features for *** tuning to control the learning process is then performed to fit the best strength models for corroded *** all the proposed data-driven models,the ANN model with three neural layers has the highest training accuracy,but also presents the largest *** SVM model provides both high training accuracy and high validation *** LR model has the best performance in terms of generalization *** models can be served as surrogate models by transfer learning with new coming data in future research,facilitating a sustainable and intelligent decision-making of corroded pipelines.