Leakage Detection Of Pipe Network Based On Wavelet Analysis And Multi-scale Correlation

For network pipelines, the distribution of pipeline arranged in a crisscross patternand the environment also embedded very complex.In the practical project condition,the application of commonly used methods in pipe leakage detection was oftenlimited by many factors of the scene.Most of the method is difficult to achieve thedesired accurate detection and positioning effect of the leakage signal.For the waveletanalysis can be performed on the signal singularity analysis and The correlationanalysis algorithm is simple and does not need to establish mathematicalmodel.Combine with these two algorithms,the paper focus on the fusion algorithm ofthe two algorithms for leak detection analysis.,this paper determinated to use thewavelet multi-scale and correlation analysis in the leakage detection algorithm of pipenetwork of infrasonic wave.The reason to choose infrasonic wave as the leakdetection object is that the wavelength of infrasonic wave is long,energy loss in thetransmission process is small,has the very good anti-interference ability and also easyto detect the location of leakage signal analysis.The main work of this paper are asfollows：Firstly,the paper describes the domestic and abroad development status and thedevelopment trend of pipeline leakage detection.Through the summary andcomparative analysis of all kinds of leak detection methods,this paper determined thewavelet multi-scale and correlation analysis in the leakage detection algorithm of pipenetwork of infrasonic wave.Then,the first chapter analyzes the signal noise reductionpretreatment,it introduces the principle of wavelet threshold denoising algorithm,infrasound signal and treat the generation mechanism, acquisition methods areintroduced, by MATLAB simulation tool, results show that the background noise ofthe threshold denoising algorithm on infrasonic wave signals have very good effect,improve the sound signal to noise ratio, do the preliminary work well for subsequentsignal analysis.Secondly,the paper introduces the leak detection algorithm based on waveletdecomposition of the infrasound pipe network. The last chapter tell the acoustic leakdetection algorithm based on detection accuracy problems left, on the basis ofanalyzing the signal threshold denoising, the leak detection algorithm of waveletdecomposition coefficients of high frequency domain transform sound based ondomain transform, the high frequency coefficients of wavelet decomposition of thegeneration, a new set of high frequency coefficients were reconstructed algorithm.Simulation results show that, the high frequency coefficients transform algorithmsobserved infrasound signal peak mutation range is more concentrated, can get timedomain signal sampling point corresponding to the mutation position accurately, and improve the accuracy of the expected pipeline leak detection.Then,the third chapter introduces leak detection algorithm of the infrasonic wavesignal based on multi-scale cross-correlation in pipeline. Compared with thecalculation error of single scale correlation analysis leakage detection method,thepaper determine to use the improved method in multi-scale correlation analysisalgorithm.Simulation analysis shows that, the cross-correlation analysis of multi-scalesignal can be accurately obtained related peak time estimation, improve the leakdetection precision of infrasound.Last but not least,the paper introduces leak detection algorithm based on thewavelet multi-scale and correlation analysis of pipe network based on infrasonic wave.Fusion of wavelet decomposition coefficients of high frequency domain transformalgorithm and multi-scale correlation analysis algorithm, the multi-scale wavelet forsignal layer correlation analysis. And the results are compared with the former twokinds of algorithm simulation results and data obtained, detection and localization offusion algorithm is more precise, and further improve the accuracy of infrasound pipenetwork leak detection and location.