从管段走向管网:管道泄漏诊断技术研究进展  

From Pipeline Segments to Pipeline Networks:Research Progress in Pipeline Leak Diagnosis Technology

作　　者：张化光[1,2] 王天彪 胡旭光 马大中[1] 刘金海[1] ZHANG Huaguang;WANG Tianbiao;HU Xuguang;MA Dazhong;LIU Jinhai(College of Information Science and Engineering,Northeastern University,Shenyang 110819,China;State Key Laboratory of Synthetical Automation for Process Industries,Northeastern University,Shenyang 110819,China)

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110819 [2]东北大学流程工业综合自动化全国重点实验室,辽宁沈阳110819

出　　处：《控制工程》2024年第6期961-972,共12页Control Engineering of China

基　　金：国家重点研发计划资助项目(2018YFA0702200);国家自然科学基金资助项目(U23B20118)。

摘　　要：管道泄漏诊断技术在保障管道系统安全运行中起着至关重要的作用。首先,介绍了管道泄漏诊断系统的结构,并指出由单一管段向复杂管网泄漏诊断的发展趋势。进一步从基于数据驱动的传统泄漏检测方法、管道泄漏信号源定位技术和基于深度学习的复杂管网泄漏检测方法3个方面进行综述,分析了不同方法的优势、局限性和适用范围。最后指出,随着管网系统的复杂度增加,传统方法的局限性逐渐显现,基于深度学习技术的复杂管网微弱泄漏诊断、多源信号融合和管网智能化的研究将成为未来的研究趋势。Pipeline leak diagnosis technology plays a crucial role in ensuring the safe operation of pipeline systems.Firstly,the structure of pipeline leak diagnosis systems is introduced,and the trend of transitioning from single pipeline segments to complex pipeline networks for leak diagnosis is highlighted.Furthermore,a review is provided on traditional leak detection methods based on data-driven approaches,pipeline leak source localization techniques,and complex pipeline network leak detection methods based on deep learning.The advantages,limitations,and applicability of different methods are analyzed.Finally,it is pointed out that with the increasing complexity of pipeline network systems,the limitations of traditional methods are gradually becoming apparent,and research on deep learning-based techniques for detecting minor leaks in complex pipeline networks,multi-source signal fusion,and pipeline network intelligence is identified as future research trends.

关 键 词：管道泄漏诊断 深度学习 复杂管网 自适应动态规划 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]