基于极限学习机的输水管网暗漏预测方法研究  

Research on Prediction Method of Hidden Leakage in Water Transmission Network Based on Extreme Learning Machine

作　　者：王小娟[1] 胡兵 马燕 刘文 WANG Xiao-juan;HU Bing;MA Yan;LIU Wen(Department of Mathematics and Physics,Xinjiang Institute of Technology,Urumqi Xinjiang 830023,China;Department of Control Engineering,Xinjiang Institute of Technology,Urumqi Xinjiang 830023,China)

机构地区：[1]新疆工程学院数理学院,新疆乌鲁木齐830023 [2]新疆工程学院控制工程学院,新疆乌鲁木齐830023

出　　处：《计算机仿真》2022年第10期506-510,共5页Computer Simulation

基　　金：2019年度新疆工程学院科研育人基金项目(2019xgy682112);国家自然基金(61962058);新疆维吾尔自治区高校科研计划自然科学项目青年项目(XJEDU2020Y043)。

摘　　要：针对输水管网暗漏难以预测的问题,提出一种基于极限学习机(Extreme Learning Machine, ELM)的输水管网暗漏预测方法。方法通过分析某校园老医务室楼用水量数据,选取用水量数据平稳的五月、六月夜间用水量数据作为数据集,建立BP神经网络、RBF神经网络、ELM神经网络预测模型,预测结果表明,ELM神经网络具有最优的预测效果,RBF神经网络预测效果相对较好,BP神经网络预测效果较差。ELM能够很好的预测输水管网的暗漏,为输水管网暗漏检测及预测提供借鉴和参考。Aiming at the problem that it is difficult to predict the hidden leakage of water transmission network, a method of prediction based on Extreme Learning Machine has been proposed. By analyzing the water consumption data of an old medical building on a campus, the data of the night water consumption with stable water consumption data in May and June were selected as the data set. The prediction models of BP neural network, RBF neural network and ELM neural network were established. The prediction results show that ELM neural network has the best prediction effect, RBF neural network has a relatively good prediction effect, and BP neural network has a poor prediction effect. ELM can well predict the hidden leakage of water delivery network, which provides a reference for the hidden leakage detection and prediction of water delivery network.

关 键 词：输水管网 暗漏 极限学习机 预测 神经网络 

分 类 号：TV213.4[水利工程—水文学及水资源]