Abstract： A recognition method for safety-detection events of oil and gas pipeline based on the wavelet energy spectrum and wavelet information entropy is studied, which is used in the distributed optical fiber oil and gas pipeline safety detection system based on the principle of Mach-Zehnder optical fiber interferometer for the safety of oil and gas pipeline. In this pre-warning system an optical cable is laid along the pipeline in the same ditch and three single mode optical fibers in the optical cable build up the distributed micro-vibrant measuring sensor. The vibration signal caused by Leakage and other abnormal events can be detected by the system in real-time. The detection signals of abnormal events were decomposed into sub signals in different frequency band by use of wavelet transform with multi-resolution analysis. Thus both frequency band energy features and wavelet information entropy of above-mentioned detection signals are extracted; the system can recognize abnormal intrusion events which have occurred along the pipeline through the wavelet energy spectrum and wavelet information entropy in real-time. Finally, the experimental data obtained at GangZao products pipeline are used to evaluate the method and the overall recognition accuracy rate achieves 98.5％, which proves the feasibility and effectiveness of this method.
2010, Vol. 29 
