管道高后果区智能识别及管理软件设计  

Design of Intelligent Identification and Management Software forHigh Consequence Area of Pipeline

作　　者：张庶鑫 张朋岗[2] 张伟[2] 熊林[2] 朱娟娟 ZHANG Shuxin;ZHANG Penggang;ZHANG Wei;XIONG Lin;ZHU Juanjuan(State Key Laboratory for Performance and Structure Safety of Petroleum Tubular Goods and Equipment Materials,CNPC Tubular Goods Research Institute,Xi′an,Shaanxi 710077,China;Oil and Gas Engineering Research Institute,PetroChina Tarim Oilfield Company,Korla,Xinjiang 841000,China;School of Aerospace Science and Technology,Xidian University,Xi′an,Shaanxi 710000,China)

机构地区：[1]中国石油集团石油管工程技术研究院,石油管材及装备材料服役行为与结构安全国家重点实验室,陕西西安710077 [2]中国石油天然气股份有限公司塔里木油田分公司,新疆库尔勒841000 [3]西安电子科技大学空间科学与技术学院,陕西西安710000

出　　处：《石油管材与仪器》2021年第5期1-7,共7页Petroleum Tubular Goods & Instruments

基　　金：“十三五”国家重点研发计划项目“原油天然气储罐及附属管道、辅助设施安全评定与风险评价预警研究”(编号:017YFC0805804)。

摘　　要：管道高后果区是指管道泄漏后可能对公众和环境造成较大不良影响的区域,因此高后果区管理是运行单位管理的重点,须定期对高后果区进行识别。常规的识别及统计方法受人为因素影响较大,导致识别过程冗长,人力与物力消耗大,识别结果查询不便、通用性差、后处理困难等问题。针对高后果区管理需求,研究了基于GIS的油气管道信息管理软件方案,并结合数据库的优势提出了一种高后果区识别改进方案;针对管道周边建筑目标检测,提出了基于ResNet-101的Faster RCNN检测方法。试验数据表明,所设计的系统架构提供了油气管道的可视化分层管理的可行方案,且高后果区识别算法可以自动识别管道沿线潜在高后果区,为高后果区识别工作提供参考,降低现场工作量,提高识别精度。The high-consequence area of the pipeline refers to the area that may have a greater adverse impact on the public and the environment after the pipeline leaks.Therefore,the management of the high-consequence area is the focus of the management of the operating unit,and the high-consequence area must be identified regularly.Conventional identification and statistical methods are greatly affected by human factors,resulting in a lengthy identification process,a large consumption of manpower and material resources,inconvenient query of identification results,poor versatility,and difficulty in post-processing.In response to the needs of high-consequence area management,this paper studies the GIS-based oil and gas pipeline information management software program and combines the advantages of the database to propose an improvement plan for the identification of high-consequence areas.For the detection of building targets around the pipeline,a Faster RCNN detection method based on ResNet-101 is proposed.Experimental result shows that the system architecture designed in this paper provides a feasible solution for visualized hierarchical management of oil and gas pipelines,and the high-consequence area identification algorithm can automatically identify potential high-consequence areas along the pipeline,which provides a reference for the identification of high-consequence areas and reduces on-site workload as well as improves recognition accuracy.

关 键 词：GIS 高后果区 目标检测 缓冲区分析 油气管道 

分 类 号：TE88[石油与天然气工程—油气储运工程]