Font Size: a A A

Study On Distributed Optical Fiber Petroleum Pipelines Safety Detection Signal Processing Technology

Posted on:2011-04-04Degree:DoctorType:Dissertation
Country:ChinaCandidate:J C ZhangFull Text:PDF
GTID:1118330338989110Subject:Precision instruments and machinery
Abstract/Summary:PDF Full Text Request
By virtue of a great many advantages, pipelines have become the principal means of oil and gas transportation. However, pipeline leakage accidents may cause loss of life and properties along with environmental pollution. Pipeline safety detection technologies have been playing an important role in protecting the security of pipeline daily transportation. However, at present the technology can only find the fault spot after the leakage, which can reduce but can't avoid the loss. Safety early-warning technology of petroleum Pipeline could carry out early detection, pre-warning and location of the acts threatening the pipeline safety. And thus, it owns great social and economic benefits.Based on the distributed optical fiber petroleum pipeline safety early-warning detection system used Mach-Zehnder optical fiber interferometer, key technologies of petroleum pipelines safety detection signal processing are studied deeply to reduce the noises from detection signals to get effective information in a complex environment, accurately and quickly identify the pipeline intrusion events in the long-distance conditions and precision fault location technology in this paper. The major study of this dissertation covers the following aspects:1. Aiming at the chaotic characteristics of the distributed optical fiber petroleum pipeline safety detection signals, the largest Lyapunov exponent is adopted as a performance index to evaluate the de-noising effect. Wavelet threshold de-noising methods and wavelet independent component analysis adapitve noise canceller de-noising methods are applied to reduce noises from detection signals acquired on site. Comparing with the signal processing results, this paper gets the best combination of mother wavelet functions and thresholds to reduce noises effectively.2. The signal feature extraction methods based on wavelet energy spectrum, wavelet information entropy and the largest Lyapunov exponent X ? {x 1 , x2 , ??? x9}are applied to extract the feature of the petroleum pipeline safety detection signals.3. Applies the BP neural network, RBF neural network and Elman neural network to recognizing the types of the detection signals along pipelines. Make a comparison of various performances among the above three networks, results demonstrate: the BP neural network shows great advantages in generalization ability and the recognition correct rate. 4. Aiming at the limitation of conventional direct cross-correlation algorithm for time delay estimation, Three location methods are proposed to improve the location precision, which are independent component analysis (ICA) time delay estimation method, wavelet packet decomposition with Multi-scale Cross-correlation time delay estimation method and order statistic correlation coefficient (OSCC) time delay estimation. Though comparison of various performances among the above three location methods, the wavelet packet decomposition with Multi-scale Cross-correlation location method are used to improve the location precision.This paper is supported by the National Natural Science Founds of China(NNSF)—Study on new method and key technology of fluid-filled pipeline network leakage detection (NO:60534050) and Science and Technology Key Projects of Petro China Company Limited—study on the pre-warning techniques based on the distributed optical fiber for the long-distance oil and gas pipelines .
Keywords/Search Tags:petroleum pipeline, distributed optical fiber sensor, signal denoising, feature extraction, pattern recognition, fault location
PDF Full Text Request
Related items