Session: DA-08-03 Developments in FFS Assessment

Paper Number: 124240

124240 - A Calculation Method of Pipeline Bending Strain Based on Imu Data of Pipeline Internal Detection 

In the process of service, long-distance oil and gas pipelines are located in a complex and changing environment, and the pipelines are prone to bending and deformation under the action of various factors, and in serious cases, wall rupture will occur, which will lead to major pipeline failure accidents. Therefore, accurately calculating the bending strain of the pipeline allows us to grasp the actual bending state of the pipeline, thus maximizing the prevention of pipeline failure due to bending deformation. In this paper, we propose a method to calculate the bending strain of pipeline based on inertial measurement unit data, which can accurately identify the bending deformation occurring in the pipeline; the change of pipeline center line position and the change of angle in the tangential direction of pipeline center line are deduced from the triaxial acceleration value and triaxial angular velocity value, so that the bending strain at each point of the pipeline can be calculated. When the bending deformation occurs in the straight pipe section, the calculated bending strain has the characteristics of the upper convex function, and the vertex of the upper convex function corresponds to the location near the maximum bending deformation of the pipe section, and the larger the bending deformation, the more obvious the characteristics of the upper convex function. Finally, this spatial feature is effectively extracted from the data. the reliability of the algorithm is verified by analyzing the calculation results of the actual pipeline detection data; the regular characteristics of the change in bending strain results are consistent with the reality, and the effective recognition rate of the elbow reaches 90%.

Presenting Author: Dong Xie China University of Petroleum(Beijing)

Presenting Author Biography: Dong Xie received the B.S. degree in Intelligence Science and Technology from China University of Petroleum (Beijing), Beijing, China, in 2021. He is currently working torwards the Ph. D in Safety Science and Engineering with China University of Petroleum (Beijing), Beijing, China. His research interests include Artificial Intelligence and Pipeline Safety.

Authors:

Dong Xie China University of Petroleum(Beijing)
Yi Shuai China University of Petroleum (Beijing)
Jian Shuai China University of Petroleum, Beijing
Fei Ren China University of Petroleum (Beijing)
Cuicui Chen PipeChina West Pipeline Company Limited
Yuanliang Jiang Sino-Pipeline International Company Limited
Haipeng Liu Sino-Pipeline International Company Limited