Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Chinese Journal of Scientific Instrument》 2014-06
Add to Favorite Get Latest Update

Identification of water pipeline leakage based on acoustic signal frequency distribution and complexity

Wen Yumei;Zhang Xueyuan;Wen Jing;Zhen Jinpeng;Wang Kai;Research Center of Sensors and Instruments,School of Opto-electronic Engineering,Chongqing University;  
In order to achieve effective leak detection of water pipeline under different conditions,this paper proposes a method to identify pipeline leak signals using the energy spectrum width parameter and approximate entropy as the feature parameters. According to the difference of the energy distribution in frequency spectrum between leak signal and noise signal,the energy spectrum width is used as feature parameter to identify the leak signals and narrow band noises wide band noises. To deal with the misjudgment caused by the fixed interference noises that have similar energy distribution in the frequency spectrum of the leak signals,the stochastic difference between these two kinds of signals is used and the approximate entropy complexity is introduced as the feature to identify the leak signals and fixed interference noises. The support vector machine is developed as a classifier,the energy spectrum width and approximate entropy of the denoised detection signal are taken as the network inputs of the support vector machine,and the pipeline leakage is identified.The detection signals in engineering application were processed with the proposed method. The processing results show that the correct leakage detection ratio is greater than 93%,and this method can effectively identify the leak signals from various non-fixed interference noises and fixed interference noises in water pipeline.
【Fund】: 国家自然科学基金(61174017)资助项目
【CateGory Index】: TU991.36;TN911.7
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved