【作者】 金文；

【导师】 陈长征；

【作者基本信息】 沈阳工业大学 ， 机械设计及理论， 2008， 博士

【摘要】 本文综述了国内外动态回转机械和静态承压装备状态监测和故障诊断技术的发展和研究状况及存在的问题,介绍了基于物元模型的可拓理论基本框架,描述了事物的物元表达形式,并将可拓理论引入动态回转机械和静态承压装备状态监测和故障诊断中,引人关联度概念对故障进行定量和定性分析,以突破经典集合理论中用单纯是、非来判断进行故障诊断的局限。用可拓理论的观点分析了设备故障特征的发散性、可扩性、相关性和共轭性,提出了可拓变换中的物元菱形思维方法,将多维物元的理念应用到故障诊断中,为较全面地描述故障特征提供了一个新思路。应用物元理论,建立了故障特征的物元模型和故障状态的物元模型,通过引入接近阈和类别阈的概念,提出了多故障诊断的最大可能性原则和最小接近阈原则及其相应的可拓故障诊断方法,定性地确定故障类别和定量地确定故障发生的可能性大小,从而给出定性、定量判断故障的依据,为多故障诊断提供了一种新的方法。利用提出的可拓诊断方法对汽轮机组的故障进行诊断分析,结果表明所提出的可拓诊断方法可以有效地对汽轮机组的多故障进行诊断。针对多特征参数故障的参数范围交叉与重叠会严重降低故障识别能力的问题,本文对影响故障识别能力的因素进行了分析研究,提出了二次特征参数的概念,认为除特征参数本身外,特征参数之间的关系可以作为二次特征参数反映故障的特征,可用于故障识别,提出了提取故障二次特征参数的比值法。重点研究了二次特征参数提取的参数差值比值法,提出了寻找最优二次特征参数的相对均方根距法和节域均值关联度法及其相应的算法准则,即最大相对均方根距准则和最小节域均值关联度准则,推导并实现了寻找最优二次特征参数的算法,得到了一种故障二次特征识别的新方法,为故障特征识别提供了一条新的途径。根据承压特种设备故障声发射特征参数的特点,提出了承压特种设备故障诊断的声发射参数比值法,建立故障特征的物元模型和故障状态的物元模型,利用提出的可拓诊断方法对承压特种设备故障进行诊断,明显提高了故障识别能力,解决了多特征参数故障在参数范围交叉与重叠情况下的故障识别与诊断问题。更多还原

【Abstract】 The internal and international progresses as well as the problems in the condition monitoring and fault diagnostics of rotating machinery and pressurized equipments are reviewed. The foundation of extension theory, which is based on matter element model, is introduced. The expressions of matter element for things are described, while the extension theory is applied to the condition monitoring and fault diagnostics of rotating machinery and pressurized equipments. In order to overcome the restriction on the fault diagnosis in classical set theory, the multiple faults are analyzed quantitatively and qualitatively with the help of the concept of correlation degree instead of the concept of yes-no. The divergence, extensibility, relativity and conjugate of fault characteristics of equipments are analyzed from the point of view of extension theory. Rhombus thought of matter elements is put forward and the idea of multiple matter elements is used to fault diagnosis, which supplies a new approach to describe the characteristics of the faults. The matter element model of the fault characteristics and fault situation are established. By introducing the concepts of the difference field and the recognition field, the biggest dependence rule and the smallest close field rule are brought forward, and a new multiple faults diagnostic method based on extension theory is deduced which can determine the type of faults qualitatively and the possibility quantitatively. The multiple faults diagnosing of a gas turbine is illustrated by using the deduced method, which shows that the method is effective on multiple faults diagnosis of the gas turbine.In order to solve the problem that the recognition ability weakens badly when the ranges of the characteristics of faults with multiple parameters superpose or overlap, the factors affecting the recognition ability to faults are investigated. It is supposed that the relations between characteristic parameters of the faults can be regarded as secondary parameters which are the characteristic parameters too. The concept of secondary characteristic parameters is put forward and the ratio method to extract the secondary characteristic parameters is deduced. The parametric difference ratio method is investigated with emphasis to extract the secondary characteristic parameters, and the relative root mean square method and correlative degree method of mean admittable field are put forward to find the best secondary characteristic parameters. The corresponding algorithms are deduced and realized. According to the characteristic of the acoustic emission from pressurized special equipments, Three-parameter-patio pethod is brought forward used to diagnose the faults of pressurized special equipments, and the recognition ability is improved obviously, solving the problem of the fault diagnosis with multiple parameters while the ranges of the characteristics of faults superpose or overlap.更多还原