【作者】 曲志刚；

【导师】 靳世久；

【作者基本信息】 天津大学 ， 精密仪器及机械， 2007， 博士

【摘要】 目前,我国国民经济的持续高速发展对能源特别是油气资源的需求越来越迫切。管道由于自身具备的诸多优势,已成为主要的油气资源运输手段。但由于种种自然或人为原因,管道泄漏事故时有发生,有时会伴随着巨大的生命财产损失和环境污染。目前,常用的油气管道安全监测装置主要通过管道输送压力、流量以及温度等参数的变化来判断泄漏是否发生。该类方法容易受输送介质特性、工艺等因素影响,且报警都在发生泄漏之后。本文在课题组前期工作基础上,对分布式光纤油气长输管道泄漏检测及预警技术展开了深入研究,并构建了一套分布式光纤油气管道泄漏检测及预警系统。该系统不仅能检测到管道泄漏并对泄漏点定位,而且能在威胁油气管道安全的异常事件发生时进行预警并对事发点定位。本文主要进行以下几方面的研究工作:1、分析了分布式光纤振动传感器机理、灵敏度和影响其灵敏度因素,同时分析了单模光纤中光波偏振态不稳对传感器的影响并比较几种解决方案。2、提出将基于小波包分解的“能量—模式”检测信号特征提取方法、基于经验模态分解(EMD)的检测信号特征提取方法、基于多尺度混沌特性分析的检测信号特征提取方法用于对管道沿线检测信号的特征提取。3、提出将径向基(RBF)神经网络和支持向量机(SVM)两种模式识别方法用于对管道沿线检测信号类型的识别。4、设计了一套实验装置,通过大量现场实验对分布式光纤油气长输管道泄漏检测及预警技术中的识别及定位性能进行验证,实验结果证明该技术对管道沿线发生的包括泄漏在内的异常事件的监测、预警、识别及定位是有效的。5、设计了一套分布式光纤油气管道泄漏检测及预警系统,在一条成品油管道上试运行。结果表明,本系统能够实时发现管道沿线异常事件并准确地对其进行识别和定位。更多还原

【Abstract】 The sustained high-speed national economic development makes the demand of energy supplies especially oil and gas resources increasingly urgent. By virtue of a great many advantages, pipelines have become the principal means of oil and gas transportation. Sometimes, however, pipeline leakage takes place due to some natural or artificial damages. Leakage accidents may cause loss of life and properties along with environmental pollution. Presently the prevalent methods to judge leakage are evaluation of parameters such as pressure in pipeline, flow rate, and temperature obtained by common oil and gas pipeline leakage detecting devices to infer whether a leakage has occurred. These methods are subject to the quality of the material transported, transportation process employed and other factors. In addition, warning of leakage exclusively arrives after the leakage.Based on the previous team work, this dissertation involves the in-depth study on the leakage detection and pre-warning technique based on distributed optical fiber for the long oil and gas pipeline. It also constructs a leakage detection and pre-warning system based on distributed optical fiber for the oil and gas pipeline. Aside from detecting and locating the leakage points, the system can also perform pre-warning monitoring and locating when the abnormal events dangerous for the oil and gas pipeline take place.The major study of this dissertation covers the following aspects:1. It elaborates the principle of distributed optical fiber vibration sensor, pressure sensitivity, and factors affecting pressure sensitivity. It also analyzes to what extent the instability of polarization state of light wave has impact on the sensor in the single-mode optical fiber, and compares several approaches to the this issue.2. It applies the“energy- pattern”signal feature extraction method based on the wavelet package decomposition, the EMD (Empirical Mode Decomposition) based signal feature extraction method, and the signal feature extraction method based on the analysis of multi-dimensional chaotic characteristics to extracting the eigenvectors of the detected signals along pipelines.3. It applies the RBF (Radial Basis Function) neural network approach and SVM (Support Vector Machine) approach to recognizing the types of the detected signals along pipelines. 4. A set of experimental devices are designed to verify the performance of the pattern recognition and locating techniques in leakage detection and pre-warning system of the distributed optical fiber oil and gas pipeline through a great many field experiments. The experiment result demonstrates that the technique has a positive effect on monitoring and pre-warning as well as identifying and locating the abnormal events including leakage along pipelines.5. A leakage detection and pre-warning system based on distributed optical fiber for the oil and gas pipeline has been accomplished, which is being operated in a product oil pipeline. The results show that the system is able to detect the abnormal events in real-time and accurately recognize the types of them and locate the places where they take place.更多还原