极限学习机在供水管网爆管定位中的应用研究  

Study on the application of extreme learning machine in the location of pipe burst in water supply network

作　　者：刘云慧 田一梅[1] 徐陶 郑国磊 彭森[1] Liu Yunhui;Tian Yimei;Xu Tao;Zheng Guolei;Peng Sen(School of Environmental Science and Engineering,Tianjin University,Tianjin 300350,China)

机构地区：[1]天津大学环境科学与工程学院,天津300350

出　　处：《给水排水》2020年第12期112-117,121,共7页Water & Wastewater Engineering

基　　金：国家重点研发计划(2016YFC0802400)。

摘　　要：针对给水管网爆管定位问题,设定爆管特征值为变量参数,将极限学习机算法与爆管事故水力模拟相结合,建立了管网爆管定位模型,并利用Matlab编程求解。将模型应用于实际管网爆管模拟,对定位方法进行了检验。分析发现压力监测点的布置、爆管管径、爆管节点的流量及其爆管特异性等因素对爆管定位精度产生了综合性影响。其中,模型的爆管定位精度与压力监测点密度及设置的均匀程度呈高度相关。结果表明基于极限学习机的爆管定位模型有较好的使用效果和应用前景。Aiming at the problem of locating the pipe burst in the water supply network,the characteristic value of the pipe burst was set as a variable parameter,and the extreme learning machine algorithm was combined with the hydraulic simulation method for establishing burst positioning model.Matlab software was used to solved the model.The model was applied to the burst simulation of actual pipe network,and the positioning method was tested.It was analyzed that the layout of the pressure monitoring points,the diameter of the burst pipe,the flow and the specificity of the burst point had a comprehensive impact on the positioning accuracy of the burst positioning accuracy.Especially,the positioning accuracy of the model had a great correlation with the placement and density of the pressure monitoring points.The result shows that the squib positioning model based on the extreme learning machine has good application effect and prospect.

关 键 词：供水管网 爆管 水力模拟 极限学习机 定位模型 

分 类 号：TU990[建筑科学—市政工程]