供水管网漏损和定位的实时三级预警模型

Real-time Three-level Warning Model for Leakage and Location of Water Supply Network

为了对供水管网的漏损和定位实现实时而精准地预报,本文建立了概率分布模型,把漏水量视为连续型随机变量,假设其服从指数分布,使用矩估计法完成参数估计,使用K-S(kolmogorov-smirnov)检验法完成假设检验,在此基础上获得了25%、50%、75%分位数,成为划分漏水程度的3个阈值,从而建立漏水程度的黄色、橙色、红色三级预警模型。以某高校供水管网91个水表的全年流量为依据,实证研究结果表明,该模型能够准确预报每个水表(管道)在各个时刻的漏水程度,为实现供水管网漏损和定位的实时预报提供新的思路和方法。

In order to achieve real-time and accurate forecasting of water supply network leakage and location,a probability distribution model is established,considering water meter leakage as a continuous type random variable,assuming it obeys exponential distribution,completing parameter estimation using distance estimation method and hypothesis testing using K-S(kolmogorov-smirnov)test method,based on which 25%,50%and 75%quartiles are obtained to become three thresholds of water meter leakage degree,thus a three-level warning model of yellow,orange and red for the degree of water meter leakage was established.Based on the annual flow rate of 91 water meters in a university water supply network,the results of the empirical study show that the model can accurately forecast the leakage degree of each water meter at each moment,which provides a new idea and method to realize the real-time forecasting of water supply network leakage and location.