基于小波变换的供水管网漏损量化方法研究

发布时间：2018-04-29 14:20

本文选题：城市供水管网 + 漏损量化　；参考：《哈尔滨工业大学》2015年硕士论文

【摘要】：漏损是城市地下供水管网面临的普遍现象,且城市漏损率呈上升趋势,远超过供水行业的指标要求。漏损量及漏损率是衡量供水管网运行状况好坏的重要指标,由于城市地下供水管网复杂、扩建速度快、漏损不可视化及供水行业管理水平限制等原因,漏损量难以准确计算。国内外现有的供水管网漏损量化方法有水平衡法、夜间最小流量法、漏损预测法、盲源分离及卡尔曼滤波法,前三种不可避免的加入了估算项,在进行控漏前不能很好的计算区域漏损程度,只能估计每月、每日漏损总量,难以得到实时漏损量数值;后两种方法利用信号处理方法进行供水管网漏损量化取得了较大的成功,但应用并不成熟。量化城市供水管网漏损,计算区域漏损率为管网控漏提供依据,预测管网经济控漏水平,模拟潜在效益;供水管网漏损得到量化,根据供水管网代表性节点的压力值,拟合压力漏损模型,指导区域进行压力控漏,建立控漏模式方案;量化城市供水管网漏损,可以改变供水管网漏损现状、节约水资源、提高供水效益。因此,研究供水管网漏损量化方法具有很大的实际意义。本文在算例管网、实验室管网采集的入口流量信号、入口压力信号及管网平均压力基础上,利用盲源分离和小波变换进行漏损量化对比研究。本课题的主要内容包括以下几个方面:1.利用EPANET建立数值模拟算例,建立漏损控制实验平台,设计管网典型供水方案,模拟不同工作状态、不同漏损数目和位置的漏损情况,采集管网流量和压力信号,为供水管网漏损量化提供数据基础;2.介绍盲源分离理论,建立盲源分离漏损量化模型,针对盲源分离幅值不确定性问题提出利用流量平衡解超定方程组解决;将小波变换方法首次应用到供水管网漏损量的计算中,建立小波变换漏损量化模型,利用相关系数选择数值模拟算例流量信号、实验室流量信号小波函数和分解层数的最佳组合,编写小波变换漏损量化程序;从理论适用性、计算漏损量精度、效果方面分析比较两种方法;3.利用盲源分离和小波变换计算算例管网、实验室管网在典型实验方案下的漏损量,拟合小波模拟漏损量、盲源分离漏损量与管网平均压力关系;比较分析两种方法量化漏损效果及压力漏损拟合效果,最终得出小波变换漏损量化效果总体优于盲源分离。
[Abstract]:Leakage is a common phenomenon in urban underground water supply networks, and the leakage rate of cities is on the rise, which far exceeds the requirements of water supply industry. Leakage loss and leakage rate are important indexes to measure the running condition of water supply network. Due to the complexity of urban underground water supply network, the rapid expansion speed, the unvisualization of leakage loss and the limitation of water supply industry management level, it is difficult to calculate the leakage amount accurately. The existing quantitative methods of leakage loss in water supply networks at home and abroad include water balance method, nocturnal minimum flow method, leakage prediction method, blind source separation method and Kalman filter method. The leakage degree of the area can not be calculated well before the leakage control is carried out, and the total amount of leakage loss can only be estimated every month and every day, so it is difficult to obtain the real time leakage amount value. The latter two methods have achieved great success in quantifying the leakage of water supply network by using signal processing method, but their application is not mature. Quantifying the leakage of the urban water supply network, calculating the leakage rate of the region to provide the basis for the pipe network to control the leakage, predicting the economic level of the pipe network controlling the leakage, simulating the potential benefit, quantifying the leakage of the water supply network, according to the pressure value of the representative node of the water supply network, Fitting the pressure leakage model, guiding the area to control the leakage, establishing the leakage control mode scheme, quantifying the leakage loss of the urban water supply network, can change the current situation of the leakage of the water supply network, save water resources, and improve the efficiency of water supply. Therefore, it is of great practical significance to study the quantitative method of leakage loss in water supply network. In this paper, on the basis of the inlet flow signal, the inlet pressure signal and the average pressure of the pipe network collected by the pipe network and laboratory pipe network, a comparative study on leakage quantification is carried out by using blind source separation and wavelet transform. The main contents of this subject include the following aspects: 1. A numerical simulation example is established by using EPANET, an experimental platform for leakage control is established, a typical water supply scheme of pipe network is designed, the leakage situation of different working state, different leakage loss number and position is simulated, and the flow and pressure signals of the pipe network are collected. For the water supply network leakage quantification provides the data basis. This paper introduces the theory of blind source separation, establishes the quantization model of blind source separation leakage, and proposes to solve the problem of uncertainty of blind source separation amplitude by using flow balance solution to overdetermined equations, and applies wavelet transform method to the calculation of leakage amount in water supply network for the first time. The leakage quantization model of wavelet transform is established, and the optimal combination of wavelet function and decomposition layer number of flow signal, laboratory flow signal and decomposition layer is selected by using correlation coefficient. Analysis and comparison of two methods in calculating the accuracy and effect of leakage loss. By using blind source separation and wavelet transform to calculate the pipe network, the leakage amount of laboratory pipe network under typical experimental scheme is simulated by wavelet, and the relationship between the leakage amount of blind source separation and the average pressure of pipe network is obtained. By comparing and analyzing the two methods to quantify the leakage loss effect and the pressure leakage fitting effect, it is concluded that the wavelet transform leakage quantization effect is better than that of blind source separation.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU991.33

【参考文献】