新一代超高清亚毫米级管道内检测技术的研发与应用  

Research and application of a new generation of ultra-high-definition inline detection technology with sub-millimeter precision

作　　者：董绍华[1,2] 田中山 赖少川 王同德 宋执武 彭东华[1] 刘冠一 魏昊天 DONG Shaohua;TIAN Zhongshan;LAI Shaochuan;WANG Tongde;SONG Zhiwu;PENG Donghua;LIU Guanyi;WEI Haotian(Pipeline Technology and Safety Research Center,China University of Petroleum(Beijing);State Key Laboratory of Petroleum Resources and Prospecting;PipeChina South China Company)

机构地区：[1]中国石油大学(北京)管道技术与安全研究中心 [2]油气资源与探测国家重点实验室 [3]国家管网集团华南分公司

出　　处：《油气储运》2022年第1期34-41,共8页Oil & Gas Storage and Transportation

基　　金：中国石油天然气集团有限公司-中国石油大学(北京)战略合作科技专项“‘一带一路’海外长输管道完整性关键技术研究与应用”,ZLZX2020-05;中国工程院重大咨询项目“油气长输管道国家治理体系战略问题研究”课题3“油气长输管道安全与应急科技支撑体系研究”,2019-ZD-37;油气资源与探测国家重点实验室自由探索课题“基于巴克豪森效应的管道应力检测技术研究”,PRP/indep-3-1812。

摘　　要：针对油气管道中普遍存在的环焊缝缺陷、类裂纹缺陷以及针孔小缺陷检测能力和识别率较低的难题,通过理论分析、建模仿真、设备研制、现场应用等环节,自主研发了新一代超高清管道漏磁内检测器。该检测器实现了探头通道间距0.6 mm、轴向采样间距1 mm,满足海量数据存储和采集要求,信号数据采集量增加15倍,并将漏磁检测、变形检测及定位检测集成到同一台检测器上,可一次性检测出各类缺陷,同时对齐腐蚀、变形以及环焊缝缺陷。开发了基于深度学习的智能化识别分析软件,建立了BP神经网络深度学习模型,实现了缺陷的一体化识别,提高了小孔腐蚀的可检测能力,可有效识别面积小于1 t×1 t(t为管道壁厚)的缺陷。该技术初步解决了当前三轴高清漏磁内检测器无法精确描述和量化小孔腐蚀缺陷、环焊缝缺陷的突出问题,对于打破国外技术垄断和提升中国管道完整性技术水平具有重大意义。(图13,参22)In view of the low detectability and identification rate of the girth weld, crack-like and pinhole defects that generally exist in the oil and gas pipelines, a new generation of ultra-high-definition pipeline magnetic flux leakage(MFL)inline detector was independently developed through theoretical analysis, modeling and simulation, equipment research and development, field application and other links. As for the detector, the probe channel spacing is up to 0.6 mm and the axial sampling spacing is 1 mm, which meet the requirements for storage and collection of massive data. In addition, the acquisition capacity of signal data is increased by 15 times. The MFL detection, deformation detection and positioning detection are integrated into the same detector, so that all kinds of defects can be detected at one time. Meanwhile,the corrosion, deformation and girth weld defects can be aligned. Moreover, the intelligent identification and analysis software based on deep learning was developed and a BP neural network based deep learning model was established,realizing the integrated identification of defects and improving the detectability of pinhole corrosion. Hence, the defects with an area less than 1 t ×1 t(where t is the wall thickness of the pipeline) can be identified effectively. In this way, the prominent problem that the pinhole corrosion and girth weld defects cannot be accurately described and quantified with the current three-axis high-definition magnetic flux leakage internal detector is solved preliminarily, which is of great significance for breaking the foreign monopolies and improving the technical level of pipeline integrity in China.(13 Figures, 22 References)

关 键 词：油气管道 内检测 超高清 海量存储 环焊缝 小孔腐蚀 智能化识别 

分 类 号：TE832[石油与天然气工程—油气储运工程]