供水管网泄漏检测定位方法及仪器

本文研究了城市供水管道泄漏形成多相湍射流引发应力波在管壁中传播机理，分析了泄漏引发的管道横振、纵振和园环振动，提出了一系列和波特征提取指标及其离散数据算法、研究了以泄漏信号应力波时和频域特征指标构造神经网络输入和矩阵，建立了能对管道运行状况进行分类的神经网络模型，以检测管道泄漏故障的发生。研究了抗干扰管网泄漏定位方法，分析了根据管网不同特征对定位公式进行修正以提高泄漏定位精度的方法，实验研究了这一

Water pipelines leak detection and location method and instrument

The mechanism of the stress wave caused by water pipilines leak trubulence injection and its propagation in the pipeline shell are researched in this paper,the longitudinal wave,transverse wave and circle wave caused by pipeline leaks are analyzed.A series of characteristic index are described to extract the information of stress wave in time domain and frequency domain. and to compress the origin data. A back propagation neural network(BPNN) with the input matrix composed of the characters of the stress wave in time domain and frequency domain is proposed to classify the running state of the pipeline, in order to detect the leaks in the fluid pipelines. A pipeline, leak interference rejection location method was proposed, and in the wave velocity correction equation the variable pipeline structure characteristic, wave impedance, pipeline diameter, pipeline fixed condition and flowing fluid characteristic, etc., were taken into account to improve pipeling leak location accuracy. The effectiveness of the new method is proved by experiments.The method is used to design an handy instrument for detection of the pipeline leakage.