The Application Study Of Pipeline Leakage Detection Based On Empirical Mode Decomposition And Blind Source Separation

Pipeline plays a much more important role in the economy construction and people’sliving. However, once the pipeline leakage happens, it will cause serious environmentalpollution，even danger to life safety. Most of the Pipeline leakage detection methods atpresent mainly research the pipeline leakage detection and location using signalprocessing method. However, the premise of this method is the leakage signal has beenacquired, actually, effective leakage signals are often mixed with a variety of interferencesignals, thus the premise of the leakage detection and location is obtaining the effectiveleakage signal accurately. In this paper, to achieve accurate detection and location ofpipeline leakage information, the stress wave techniques in the high-pressure gas pipelineleakage detection are studied, with the blind source separation and empirical modedecomposition collecting and processing the stress wave. This paper will focus on thefollowing works:Firstly, on the basis of analyzing the development and research statuses of detectiontechnology method at home and abroad in pipeline leakage, the problem of extractingeffective leakage signal was discussed. Aiming at the mathematical models of blindsource separation under condition of different mixing methods, the instantaneous linearmixed model was built, and the overdetermined and positive definite blind sourceseparation methods were researched. On the basis of the evaluation, the experimentalstudied on FastICA algorithm of speech signal is tested, and the simulation results is putin the end.Secondly, an underdetermined blind source separation method based on empiricalmode decomposition(EMD) is presented. The mixed signals collected from severalsensors are decomposed by EMD, by obtaining the kurtosis characteristics of the signaland the remodeling and restructuring the observed signal, so that the signal dimensions areincreased. It solves the underdetermined blind sources separation. The experimentalresults show that this method can extract leakage signal effectively.Finally, aiming at the problem of traditional correlation time delay estimation method, time delay estimation algorithms based on generalized cross correlation and LMS adaptivealgorithm are raised, it improves the position error affected by the signal noise ratio reducewhich is caused by noise. The experimental results show that this method can improve theaccuracy of time delay under low signal noise ratio effectively.