故障轴承振动特性分析与典型故障诊断

发布时间：2019-03-03 00:23

【摘要】：滚动轴承是各种旋转机械中应用较广泛且易损坏的机械零部件之一，它的运行状态往往直接影响着整台机器的性能。对滚动轴承进行可靠的故障诊断能及时地发现故障，能有效地避免机器的损坏和人员伤亡等重大事故。特别在轧钢机械中，滚动轴承更是特别重要的零件，故研究滚动轴承故障诊断具有现实的意义。利用电火花破坏滚动轴承用以模拟轴承故障，采用模拟故障代替实际故障进行试验研究。搭建滚动轴承故障实验平台并在平台上采集振动信号用于试验研究。本文首先分析了滚动轴承的故障机理和振动特性，并总结了滚动轴承局部损伤类故障振动信号的周期性冲击和幅值调制的特性；然后为有效地提取滚动轴承故障特征，提出了一种基于经验模态分解法和独立分量分析法相结合的故障特征提取的方法，并应用到实际的滚动轴承实验数据处理和分析中，，表明能准确地提取滚动轴承的故障特征；其次，对支持向量机的理论做了系统研究，提出用基于交叉验证的遗传算法对支持向量机中惩罚参数C和高斯核参数γ进行寻优，确保了分类模型的最优性；最后，提出了利用分离矩阵的奇异值作为滚动轴承故障特征向量的方法，通过实际滚动轴承振动实验并采集数据，验证了该方法能较好的鉴别故障类型。采用MATLAB和LabVIEW相结合的方法完成了对滚动轴承智能故障诊断系统平台的开发。该平台界面友好，易于操作；通过在滚动轴承实验台上模拟滚动轴承的典型故障的振动试验，结果证明该诊断系统性能良好。
[Abstract]:Rolling bearing is one of the mechanical parts which are widely used and easy to be damaged in various rotating machinery. Its running state often directly affects the performance of the whole machine. Reliable fault diagnosis of rolling bearings can find faults in time, and can effectively avoid serious accidents such as machine damage and casualties. Especially in rolling machinery, rolling bearing is a very important part, so it is of practical significance to study the fault diagnosis of rolling bearing. Using the electric spark to destroy the rolling bearing is used to simulate the bearing fault, and the simulation fault is used to replace the actual fault to carry on the experiment research. A rolling bearing fault test platform is set up and vibration signals are collected on the platform for experimental research. In this paper, the fault mechanism and vibration characteristics of rolling bearings are analyzed firstly, and the periodic shock and amplitude modulation characteristics of local damage vibration signals of rolling bearings are summarized. Then, in order to extract the fault feature of rolling bearing effectively, a fault feature extraction method based on the combination of empirical mode decomposition method and independent component analysis method is proposed, which is applied to the practical data processing and analysis of rolling bearing experiment. It is shown that the fault characteristics of rolling bearing can be extracted accurately. Secondly, the theory of support vector machine is studied systematically, and the genetic algorithm based on cross-validation is proposed to optimize the penalty parameter C and Gao Si kernel parameter 纬 in support vector machine, which ensures the optimality of classification model. Finally, the singular value of the separation matrix is used as the fault characteristic vector of rolling bearing. Through the vibration experiment of the rolling bearing and the collection of data, the method is proved to be effective in identifying the fault types. The intelligent fault diagnosis system platform of rolling bearing is developed by means of MATLAB and LabVIEW. The platform has a friendly interface and is easy to operate, and the performance of the diagnosis system is proved to be good by simulating the vibration tests of typical faults of rolling bearings on the rolling bearing test stand.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH165.3

【参考文献】