检漏
泄漏
测光模式
泄漏(经济)
管道运输
管道(软件)
分离(微生物学)
计算机科学
阶段(地层学)
可靠性工程
实时计算
数据挖掘
工程类
环境工程
宏观经济学
古生物学
经济
微生物学
程序设计语言
生物
机械工程
作者
Xiaoting Wang,Junyu Li,Shuming Liu,Xipeng Yu,Ziqing Ma
标识
DOI:10.1061/(asce)wr.1943-5452.0001558
摘要
Hydraulic accidents or abnormal situations, also known as leakages, cause not only water losses but also service interruptions and other negative effects. To facilitate the rapid response of water utilities and reduce water losses caused by undiscovered leakages, a timely, efficient, and accurate detection method is required. This paper describes a novel framework for multiple overlapping leak detection and isolation in the water distribution system. The proposed method has three stages: an estimation stage to extract leak-induced data from the huge monitoring data, an identification stage to detect the presence of overlapping leaks from multiscale difference series, and a localization stage to pinpoint the leak hotspots. The proposed method detects seven leakages in real-life networks and localizes 16 overlapping leakages with an economic score of €167,981 in L-town 2019. Comparisons show that the framework can accurately detect it in a fast response time and effectively localize the leaking pipeline with good applicability.
科研通智能强力驱动
Strongly Powered by AbleSci AI