摘要：The In this paper, nonlinear analysis is used to study the possibility of chaotic behavior of pipeline pressure signal. Based on real-time data of pipeline leak monitoring system, oil pipeline pressure signals have been verified to be chao system by analyzing the chaotic characteristic of pressure time series. Six typical measured data of pipeline pressure are selected and reconstructed to the higher phase space. And then the largest Lyapunov exponent of each data is calculated to test and verify the chaotic characteristics of the pressure signal. For pipeline leakage fault diagnosis, the approximate entropy (ApEn) has been applied to extract the nonlinear and chaotic characteristics. By calculating the ApEn of normal, operation and leakage signals, the results indicate that the value ranges of three kinds of signals are above 0.35, below 0.025, and from 0.025 to 0.35. And the identification rate of pipeline leakage has been reached 90.0% only based on ApEn, which provide more effective basis for the classification and identification of chaotic characteristics of the pressure signal.
