摘要
随着数据监测成本的不断下降、管网模型的普遍应用,基于数据与模型的爆管检测及定位技术受到广泛关注,但背景噪音往往导致上述方法无法准确定位爆管。针对该问题,提出基于耦合统计及模型驱动的供水管网爆管定位方法。首先利用加权最小二乘法计算延时状态下供水管网中各节点的爆管残差,然后采用耦合统计算法对爆管残差信号进行卡尔曼滤波与累积和运算进而定位爆管。采用简化管网阐明了耦合统计算法的基本原理与计算步骤,进一步利用某园区小型管网验证该算法的可行性。结果表明,该方法能从背景噪音中识别并放大爆管信号、降低爆管定位误报率,并能同时预警与定位爆管。
With the continuous decrease of data monitoring cost and wide application of pipe network model,the detection and location technology of pipe burst based on data and model has been widely concerned.However,background noise often leads to the failure of the above methods to accurately locate the pipe burst.In order to solve this problem,a burst pipe location method of water supply pipe network based on coupling statistics and model driven was proposed.Firstly,the burst pipe residual of water supply pipe network under delay state was calculated by using weighted least square method,and then the coupling statistical algorithm was used to carry out Kalman filtering and cumulative sum operation on the residual signal of the pipe burst to locate the pipe burst sites.The basic principle and calculation steps of the coupling statistical algorithm were illustrated by using the simplified pipe network,and the feasibility of the algorithm was further verified by using a pipe network in a park.The proposed method could identify and amplify the burst signal from the background noise,reduce the false alarm rate of the burst location,and simultaneously warn and locate the burst sites.
作者
卢慢
杜坤
宋志刚
周明
董云程
黄乐烽
LU Man;DU Kun;SONG Zhi-gang;ZHOU Ming;DONG Yun-cheng;HUANG Le-feng(Faculty of Civil Engineering and Architecture,Kunming University of Science and Technology,Kunming 650500,China)
出处
《中国给水排水》
CAS
CSCD
北大核心
2021年第9期46-50,共5页
China Water & Wastewater
基金
云南省重点领域科技计划项目(202003AC100001)
国家自然科学基金资助项目(51608242)。
关键词
供水管网
爆管定位
背景噪音
模型驱动
耦合统计
water supply pipe network
burst location
background noise
model driven
coupling statistic
