关键词: 光纤, 油气管道, 安全预警系统, 经验模态分解法, 多尺度混沌特性分析法, 固有模态

Abstract: A multi-scale chaotic characteristic analysis method was proposed based on the empirical mode decomposition(EMD) for analyzing the detected signals in oil-gas pipeline pre-warning system.The pre-warning system was based on the principles of Mach-Zehnder optical fiber interferometer.An optical cable laid along the pipeline in the same ditch and three single mode optical fibers in the optical cable were made up the distributed micro-vibrant measuring sensor to detect the leakage and other abnormal events along the pipeline in realtime.The vibration signals of abnormal events were decomposed by the EMD method.The components of intrinsic mode function(IMF) obtained from the EMD were analyzed by using the multi-scale chaotic characteristic analysis method,and the correlation dimensions under different scales were calculated.This method can characterize the chaotic traits of the detected vibration signals more clearly.The field signal test proved that this method was more effective in recognizing the typical abnormal events.

Key words: optical fiber, oil-gas pipeline, safety pre-warning system, empirical mode decomposition(EMD) method, multi-scale chaotic characteristic analysis, intrinsic mode function(IMF)