基于信号处理的齿轮箱故障诊断方法研究

发布时间：2018-04-24 00:37

本文选题：齿轮箱故障诊断 + 经验模态分解　；参考：《电子科技大学》2017年硕士论文

【摘要】：齿轮箱常常运行在恶劣地环境下,在运行过程中很容易出现故障。齿轮箱一旦出现故障,轻则引起生产中断与经济损失,重则会导致严重人员伤亡。振动信号分析法是齿轮箱故障诊断最重要方法之一。在正常状态运行时,齿轮箱的振动信号一般表现为平稳性的特点,而当出现故障时,其振动信号通常表现为非线性与非平稳的形式。因此,需要使用时频分析法分析其振动信号。传统的时频分析法虽然能用于分析处理齿轮故障振动信号,但是这些方法都存在较大的局限性—非自适应性。自适应时频分析法能够满足自适应的要求。经验模态分解法、局部均值分解和极值点对称模态分解法是目前最主要的自适应时频分析法。本论文研究了主流自适应时频分析方法在故障诊断中的应用,并对它们存在的问题提出相应的改进,并将其用于齿轮箱中常用的零件的故障诊断。本论文的主要研究工作有:(1)介绍了齿轮箱主要零件齿轮的失效的形式及原因、振动机理以及故障时振动信号的模型。(2)介绍了经验模态分解法的原理及存在的主要的缺陷。针对经验模态分解法现有筛分准则存在的问题,提出了一种型号筛分终止准则。存在针对经验模态分解存在的模态混淆现象这一缺陷,提出了解析经验模态分解法。对仿真信号进行了分解,验证了该方法的有效性。(3)介绍了局部均值分解法的原理。针对局部分解存在的模态混淆现象,提出了小波局部分解法。采用小波局部分解法与总体平均局部均值法分别对仿真信号与转子碰摩故障信号进行分解,最终结果表明,本文提出的小波局部分解方法能够用于齿轮箱的转轴故障诊断。而且,与总体平均局部均值法相比,该方法具有运行效率较高,分解的时间更短,信号分解准确性更好等优点。(4)极值点对称模态分解是一种新的自适应信号时频分析法,该方法目前尚未应用于机械故障诊断中。本论文将极值点对称模态分解法与能量算子解调法结合起来,用于分析齿轮断齿故障的振动信号,从而实施对齿轮断齿故障的诊断。
[Abstract]:The gearbox often runs in bad environment, and it is easy to break down during operation. Once the gearbox fails, the light will cause the production interruption and economic loss, and the heavy will lead to serious casualties. Vibration signal analysis is one of the most important methods for gearbox fault diagnosis. In normal operation, the vibration signal of the gearbox usually shows the characteristics of stationarity, but when the fault occurs, the vibration signal of the gearbox usually shows the form of nonlinearity and non-stationarity. Therefore, it is necessary to use time-frequency analysis method to analyze its vibration signal. Although the traditional time-frequency analysis method can be used to analyze and process the vibration signal of gear fault, these methods have great limitation-non-adaptive. Adaptive time-frequency analysis can meet the requirements of adaptive. The empirical mode decomposition method, the local mean decomposition method and the extreme point symmetric mode decomposition method are the most important adaptive time-frequency analysis methods. In this paper, the application of mainstream adaptive time-frequency analysis method in fault diagnosis is studied, and the corresponding improvement of their existing problems is put forward, and applied to the fault diagnosis of common parts in gearbox. The main research work of this paper is: (1) introducing the form and reason of gear failure, vibration mechanism and vibration signal model. (2) introducing the principle of empirical mode decomposition method and its main defects. In order to solve the problems existing in the existing screening criteria of empirical mode decomposition (EMD), a model sieving termination criterion is proposed. In order to solve the problem of modal confusion in EMD, an analytical empirical mode decomposition (EMD) method is proposed. The simulation signal is decomposed and the validity of the method is verified. The principle of the local mean decomposition method is introduced. A wavelet local decomposition method is proposed to solve the modal confusion in local decomposition. The wavelet local decomposition method and the population average local mean method are used to decompose the simulated signal and the rotor rubbing fault signal respectively. The final results show that the wavelet partial decomposition method proposed in this paper can be used to diagnose the rotating shaft fault of the gearbox. In addition, compared with the average local mean method, this method has the advantages of higher running efficiency, shorter decomposition time, better signal decomposition accuracy, etc.) extreme point symmetric mode decomposition is a new adaptive signal time-frequency analysis method. At present, this method has not been applied to mechanical fault diagnosis. In this paper, the extreme point symmetric mode decomposition method and the energy operator demodulation method are combined to analyze the vibration signal of gear broken tooth fault, so as to carry out the diagnosis of gear broken tooth fault.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TH132.41

【参考文献】