【作者】 刘显德；

【导师】 刘扬； 许少华；

【作者基本信息】 东北石油大学 ， 油气储运工程， 2010， 博士

【摘要】 管道运输是石油、天然气最为经济合理的运输方式。随着油气管道的大量铺设和管道服役时间的增长,管道失效事故屡有发生,给人民生命财产带来重大损失。影响管道失效的因素众多,普遍具有随机性、模糊性和不完整性等特点,传统诊断方法对管道失效模式的分析常常存在不适应性。智能信息处理理论和技术是近几年在各工程领域和科学研究中得到广泛研究和应用的人工智能方法,其相关模型由于具有高度非线性映射能力、大规模并行分布处理和良好的自适应学习机制,很适合求解传统模式识别和预测方法难以建模解决的问题。因此,将智能信息处理方法和技术应用于管道失效模式诊断问题的研究在机制上具有很好的适应性。论文主要针对管道失效模式诊断中的若干典型问题,进行管道失效模式智能诊断理论和应用技术研究。将模式识别和动态预测领域中普遍采用的人工神经网络技术与诊断理论、模式识别、模糊逻辑和系统仿真等方法相结合,构建适合管道失效模式分析的智能诊断方法和模型,并进行求解算法和应用技术研究。在智能诊断方法和模型研究方面,论文在对油气管道已有失效模式分析和故障诊断建模技术研究基础上,归纳出三类管道失效模式诊断问题:数值型模式诊断、含模糊信息模式诊断和动态模式诊断,并构造不同的智能模型以实现上述不同问题的求解。针对数值型模式诊断问题,构建了自适应确定BP网络结构的方法和实现机制,并应用于具有较为完整测试数据的含缺陷压力管道失效模式诊断;针对含有模糊信息的失效模式诊断问题,考虑已知条件和结果之间无明确因果关系及各环境因素对结果影响的重要程度不同,在传统模糊神经网络基础上建立了加权模糊推理网络,较好解决了腐蚀数据中的模糊性信息对管道腐蚀程度的影响;对于动态模式诊断问题,将过程神经网络和径向基函数神经网络相结合,提出了一种径向基过程神经元网络的概念和模型,模型融合了过程神经网络可表达动态过程效应累积和径向基网络非线性函数逼近能力强的优势,对预测管道腐蚀速率随时间非线性变化问题具有很好的适应性。同时,针对过程变量趋势预测,将传统支持向量回归机的构造思路和方法推广到时变函数空间,建立了一种过程支持向量回归机,该模型可较好地解决动态系统时间预测问题。在应用技术研究方面,给出了智能诊断模型在一些典型管道失效模式诊断问题中的应用方法和求解过程。主要包括管道泄漏诊断、管道腐蚀失效模式诊断、管道腐蚀速率预测、含缺陷压力管道失效模式诊断以及管道防腐保温层故障诊断分析等,并获得了较好的应用结果。论文针对管道失效模式诊断中的若干典型问题,建立了相关的智能诊断模型和方法,并进行了实际应用研究。这对于油气管道失效事故分析和管道运行完整性评价提供了一种科学方法和手段,可为管网进行风险性评估与运营决策提供科学依据,具有重要的实际意义和应用前景。更多还原

【Abstract】 Pipeline transportation is the most economical and reasonable transport mode of oil and natural gas. With the large number of oil and gas pipeline laying and the growth of service time, the failure incidents of pipeline happened frequently, which brought great losses for peoples’life and property. There are many factors influencing pipeline failure. Some of them are random, fuzzy, incomplete and other characteristics, traditional diagnostic methods are often not adaptive for pipeline failure mode analysis. Intelligent information processing theory and technology is an artificial intelligence method which is researched and applied extensively in various engineering fields and science research in recent years, because its correlation model has highly nonlinear mapping capability, large-scale parallel processing and good adaptive learning mechanism, it is very suitable for solving the problems which the traditional pattern recognition and prediction methods are difficult to model. Therefore, it has good adaptability for the Intelligent information processing method and technology which are applied in the pipeline failure mode diagnosis.In view of some typical issues of pipeline failure mode diagnosis, the paper mainly researches the pipeline failure mode intelligent diagnosis theory and application technology. Combines the Artificial Neural Network theory and diagnosis theory, pattern recognition, fuzzy logic with system simulation methods which are widely used in pattern recognition and dynamic forecasting field, construt the intelligent diagnosis technology and model which are suitful for pipeline failure mode analysis, and carry on solution algorithm and application technology research.On the side of intelligent diagnosis method and model research, the paper summarize three types of pipeline failure mode diagnosis problems, they are numerical mode diagnosis, fuzzy information mode diagnosis and dynamic mode diagnosis on the basis of analyzing gas pipeline failure modes and fault diagnosis modeling technology, and construct different intelligent model to complete the solution of above-mentioned different problems.Aimed to the numerical mode diagnosis problem, we construct the adaptive method used to define BP networks structure and completing mechanism, and apply it into the failure modes of pressure pipes with defects;To the fuzzy information mode diagnosis problem, considering the unclear relationship between conditions and results and the importance degree of conditions impacting on results, a weighted fuzzy reasoning networks is constructed based on traditional fuzzy neural networks. It solves the fuzzy information of corrosion data impacting on pipeline corrosion degree better. To the dynamic mode diagnosis problem, process neural networks is combined with RBF neural networks, and the concept and model of RBF process neural networks are introduced, the model integrates the advantages of process neural networks which can express the cumulative effect of the dynamic process and the RBF networks nonlinear function has strong approximation capabilities, it has a good adaptability for the prediction problem of pipeline corrosion rate changing nonlinearly with time. At the same time, in view of the problem of process variables trend prediction, the structural ideas and methods of traditional support vector regression machines are extended to time-varying function space. We establish a process support vector regression machines, the model can solve the time prediction problem of dynamic system better.On the side of application technology, the paper gave the application methods and solution process of some typical pipeline failure mode diagnosis problems using intelligent diagnosis model. These problems include pipeline leak diagnosis, pipeline corrosion failure mode diagnosis, pipeline corrosion rate prediction, pipeline failure mode diagnosis with defective pressure and pipelines insulation fault diagnosis analysis,etc, and all of them get the better application results.The paper establishes the related intelligent diagnostic model and methods contrary to some typical problems of pipeline failure mode diagnosis, and carries on the practical application research. It provides a kind of scientific method for the oil-gas pipeline failure accident analysis and the integrity assessment of pipeline running, can provide scientific basis for risk assessment and management decision-making of pipelines, and it has important practical significance and application prospects.更多还原