故障征兆

• The research work presented a large quantity of debris characteristics parameters , and especially made a thorough study on the characteristic description of sediment chain graph ; meanwhile , the sensibility , differentiation and information redundance analyses of the characteristic parameters also supplies the quantitative indexes for the filtration and optimization of the debris characteristic parameters ; in addition , the debris fusion decision recognition method based on the proof fusion theory and the comprehensive debris recognition flow provide reliable recognition arithmetic for debris recognition ; and lastly , the fault fusion diagnosis judging method based on ferro - graph and spectral analysis provides the basic diagnosis method in theory for multi - fault premonitory diagnosis system of aero - engine
  本文研究工作提出了大量磨粒特征参数，尤其深入研究了沉积链谱片的特征描述问题；同时，特征参数的敏感性、区分度与冗余度分析为磨粒特征参数的筛选优化提供了量化指标；另外，基于证据融合理论的磨粒统计融合决策识别方法以及磨粒综合识别流程为磨粒识别提供了可靠的识别算法；最后，基于光谱和铁谱信息的磨损故障融合诊断决策方法为发动机多故障征兆综合诊断系统提供了基本的诊断理论手段。
• 2 . study on knowledge - based fault diagnosis system systemeticly to find the causes of faults , maily include the following aspects : representation of symptoms and experience knowledge ; measurement of fuzzy symptoms ; inference algorithm of the inaccuracy diagnosis ; interface of acquisition of evidence and management of knowledgebase . 3
  以详细诊断故障原因为目的，系统的研究基于知识的故障诊断系统，主要包括：故障征兆知识及经验知识的形式化表示；故障征兆知识的量化处理；不确定诊断问题的求解算法；获得诊断证据的接口及知识库管理技术。
• In order to solve the problem how to get expert knowledge and how to detect fault , a new settlement is presented in this paper , which is based on simulation . in the beginning of this paper , the writer introduces some conception of conventional expert systems , and analyzes its localization or disadvantage . according to the special of electronic devices , artificial intelligent fault detecting system of a certain radar based on simulation is put forward ; the paper gives the detail of the top bottom design for an example , and gives the solution for remote diagnosing in internet
  本文主要试图解决这么几个问题：面向故障诊断进行仿真建模；对仿真模型进行故障注入，获取故障模式，从而获取系统全面的专家知识；采用模糊神经网络对这些知识进行规约和抽象；采用模糊聚类算法对专家知识进行诊断现场的融合；采用虚拟仪表获取信号数据，运用小波变换等提取故障征兆，实现故障的诊断和预测；基于仿真模型构建嵌套推理算法，实现模拟故障排除训练和考核，分析岗位人员对原理的掌握情况。
• 3 . present corresponding database representation method for diagnostic knowledge : aimed at knowledge based on decision trees , use the parent node representation adding node state column and fault column to implement the database storage for diagnostic knowledge ; while for knowledge based on cases , the fault symptom matrix is available . 4
  诊断知识的表示方面，提出了相应的数据库表示方法：对于基于决策树的知识，用增加状态数据域和结点状态列以及故障列的双亲结点表示法，实现诊断知识的数据库存储：对于基于案例表示的诊断知识，以故障征兆矩阵的形式存入数据库。
• Chart of axes track has the important malfunction information , it is very important in diagnostics of malfunction . in practice , we can obtain earlier stage malfunction sign of the generator that comes from the chart of axes track , and it would have instructive function to preventing deterioration and excluding malfunction
  发电机组的转轴的轴心轨迹形状是重要的故障征兆,在机组故障诊断中起着重要作用。在实际操作中,根据轴心轨迹图还可以得出故障的前期征兆,对防止故障的恶化和排除故障具有指导作用。
• First , the fault type is identified by rough - set , then the neural networks determines the fault elements by the sampling voltage values . with regarding to the other method , the rough - set is looked as the pre - system of neural network . . the fault information as well as the sampling voltage values are simplified by rough - set , and the outcome of this processing is the input of neural network .
  整体结合是利用粗糙集对故障信息中样本的所有故障征兆进行数据处理,通过知识约简,删除多余的征兆属性,简化知识表达空间维数,简化以后的样本数据作为神经网络的训练样本,构成完整的粗糙集-神经网络故障诊断方法，将粗糙集与神经网络相结合,简化了神经网络结构,从而达到提高诊断速度的目的。
• Secondly , introducing the image analyzing technology with reference to the disadvantages of the traditional ferr - graph analysis technology , and with the combination of characteristic parameter optimizing filtration so as to raise a description method of debris micro - morphologic character . thirdly , with the application of mode recognition method , completing the process of debris auto - recognition based on the collected information of the debris configuration characteristics ; and conducting the diagnosis on the aero - engine wear faults according to the theory of particle tribology . fourthly , introducing information fusion technology to solve the problem that a single method can not collect enough fault premonitory information to conduct the wear fault diagnosis , hence to conduct the research and exploration in the field of comprehensive diagnosis on the aero - engine ' s multi - fault premonitory information
  本文的研究工作主要包括以下五个部分：首先，介绍航空发动机常见的磨损故障类型，研究磨损故障的失效机理，分析磨粒的产生机理、分类以及形态特征：其次，针对传统铁谱分析技术的缺点，引入图像分析技术，再结合特征参数优化筛选，形成基于图像的磨粒显微形态学特征描述方法：然后，基于提取到的磨粒形态特征信息，应用模式识别方法完成磨粒自动识别，并根据颗粒摩擦学的基本原理进行航空发动机磨损故障的诊断与定位：再后，鉴于单一方法不能提取足够的故障征兆信息进行磨损故障诊断，本文引入信息融合技术，开展航空发动机多故障征兆信息综合诊断方法的研究与探索；最后，基于航空发动机滑油光谱分析与铁谱分析数据，应用时序模型、灰色模型以及组合模型进行磨损故障的预测方法研究。
• And an intelligent fault detecting system is established . starting from analyzing the possible faults existed ; relations between the fault patterns of the wall - climbing robot and the characteristic signals are established . so the problem of fault detection of the robot is in essence a problem of pattern recognition
  主要内容如下：对壁面机器人的可能故障进行分析，总结系统可能的故障类型，并选择体现不同故障类型的特征信号，构建系统的故障空间和故障征兆空间，建立了两者之间的映射关系，并在此基础上制定了故障诊断方案。
• In this paper , overall design philosophy and measure while diagonose the prefabricated substation using ann theory are defined , including the definition of fuzzy expression method for fault symptoms , the definition of typical fault collection and typical fault sign collection , the definition of the format of the learning sample and test sample , and the definition of fault diagnosis model formed in coordination by multi ann whose diagnosis principle are also described . a practical software using visual c + + 6 . 0 and access2000 as developing instrument are developed on the basis of diagnosis principle put forward by this paper
  本文确定了应用神经网络理论对箱式变电站进行故障诊断的总体设计思想和步骤：确定了监测数据的预处理模糊化方法；建立了箱式变电站典型故障集和典型故障征兆集；确定了学习样本的格式，完成了学习样本的生成；确定了神经网络结构和参数，并对学习样本应用本文的学习算法进行了学习训练，使误差控制在给定范围内；以集散监测诊断系统的思想，提出了由多个神经网络协同构成的多神经网络故障诊断模型，并论述了其诊断原理。