A Study On The Method Of Pipeline Leakage Signal Feature Extraction

Oil, gas and other energy in people’s life and economic activity plays a very important role. The use of petrochemical products have occupied most of the energy market, pipeline as petrochemical products main means of transport also has an important position. Oil or gas pipeline generally distance is longer, but also there are many factors affecting pipeline safety operation, for example, corrosion and ageing, man-made destruction, not to force majeure factors, etc. How to reduce resource depletion and to prevent oil, gas, because has the characteristics of inflammable, explosive caused the accident is particularly important. Therefore, the real-time detection and positioning of pipeline research has been more and more attention.Pipeline exist leakage, leakage of medium eruption outward form sound waves, aiming at the existing problem of gas pipeline now, pipe sound production principle, the sound propagation in the pipeline, determine whether pipeline leakage, and the calculation method of pipeline leak. First using LabVIEW software of real-time data acquisition and acquisition interface card, and then the collected signal using wavelet transform to get preprocessing signal wavelet noise reduction processing, secondly, respectively, using wavelet transform and Hilbert Huang transform to processing, preprocessing signal extraction pipeline leakage signal, the signal to make use of the Hilbert- Huang transform the EMD(Empirical Mode Decomposition) decomposition, get different IMF component of original data, verify the Hilbert- Huang transform in pipeline leak detection application feasibility. Last layers of wavelet decomposition with the cross-correlation method in combination with each other, the sound signal under wavelet multi-scale cross-correlation analysis step by step, wavelet transform and Hilbert- Huang transform position error were analyzed, summarized their use value.Through theoretical and experimental verification, the proposed pipeline leakage feature extraction method with high feasibility and reliability of leak detection applications for pipeline leakage feature extraction provides an effective method.