Calculation of the Water Distribution Networks Based on the Theory of Slow Transient Flow

Urban water supply network is a modern urban survival and development of the infrastructure of a city,and its normal running conditions have important significance. The actual hydraulic process in the variableload water distribution networks can be treated as the slow transient flow which belongs to the unsteady flow. This paper analyzes the multi-loops network slow transient model based on graph theory,and the link flow matrix is treated as the variables of the discrete solution model to simulate the process of the slow transient flow in the network. With the simulation of hydraulic regime in an actual pipe network,the changing laws of the flow in the pipes,nodal hydraulic heads and other hydraulic factors with the passage of time are obtained. Since the transient processes offer much more information than a steady process,the slow transient theory is not only practical on analyzing the hydraulic condition of the network,but also on identifying hydraulic resistance coefficients of pipes and detecting the leakage in networks.

基于图论的环状管网慢变流的计算研究

管网负荷变动的水力过程可以抽象成非恒定流中的慢变流进行研究,这是大量管网运行中的现实现象.对基于图论的环状管网慢变流模型进行分析,利用矩阵论导出了采用链支流量矩阵计算慢变流过程的解析表达式,并在此基础上得到了适于数值模拟的离散求解模型.通过实际算例计算了管段流量和节点水头等水力要素随时间的演化过程.慢变流理论不仅对分析管网系统的水力工况有很大作用,还由于其时间过程所包含的丰富信息使其在管网的水力摩阻识别及漏失故障诊断中具有重要意义.由于引入链支流量矩阵作为因变量,在很大程度上简化了定量计算,将会非常有利于对上述系统识别问题的搜索求解.

基于慢变流理论的管网泄漏检测模型的建立

提出一种基于慢变流理论的给水管网泄漏诊断与定位模型,即首先对管网施加随时间缓慢变化的压力激励,如果管网中存在泄漏,随着压力的不同泄漏程度也不一样,进而影响到管网中其他节点压力和管段流量等参数的变化;依据系统辨识反问题分析方法,实时监测管网监测点处的状态参数,运用贝叶斯算法对模型的预测参数——泄漏位置和泄漏量的概率值大小进行寻优,寻找模型计算值与测量值之间吻合程度最好的预测参数,得到最可能的泄漏位置与泄漏量。通过对算例管网的计算结果表明,采用该模型能够对漏失点和漏失量进行满足概率要求的准确诊断。

环状热网泄漏检测模型研究

基于图论理论、变流理论及贝叶斯算法建立环状热网泄漏工况水力模型、预测误差模型。在模拟管网泄漏工况时,将泄漏归结在节点处,将泄漏流量视为该节点流量的一部分,泄漏程度用泄漏面积反映。将由热网泄漏工况水力模型得到的热网泄漏数据库数据作为测量数据,将水力模型的模拟结果作为预测数据,采用预测误差模型计算预测数据的最优估计概率,指导确定泄漏位置。