基于非平稳信号分析的滚动轴承故障诊断研究

发布时间：2018-04-11 00:28

本文选题：滚动轴承 + 局域均值分解　；参考：《燕山大学》2015年硕士论文

【摘要】：滚动轴承广泛应用于工业生产中,其作为机械设备的核心部件,运行状态直接影响到机械设备的可靠性及稳定性。因此,对滚动轴承进行故障诊断对于机械设备的运行维护具有重要意义。故障信息的特征提取是轴承故障诊断的关键,本文针对滚动轴承故障诊断,运用局域均值分解(Local Mean Decomposition,LMD)、形态滤波和近似熵理论,分别从信号滤波去噪和信号序列复杂度的角度出发,对滚动轴承振动信号提取方法进行了实验研究,为滚动轴承故障诊断的特征提取提供了理论依据。论文的主要研究工作如下:(1)针对滚动轴承振动信号的非平稳性特点,以及实际故障特征信号难以提取的问题,研究了局域均值分解方法,该方法能将复杂的非平稳信号分解成一系列调幅调频函数,实现信号中不同调制频率成分的分离,能有效的分离出故障成分,可应用到实际滚动轴承振动信号的分解中。(2)针对实际轴承振动信号噪声干扰严重的问题,研究了形态滤波算法在振动信号处理中的应用,并研究了基于信号极值点确定形态滤波结构元素长度的自适应形态滤波方法,该方法能在保持原信号面貌的基础上,最大程度的抑制冲击脉冲噪声的影响,并将滤波后的信号进行局域均值分解提取故障特征信息,仿真实验表明将形态滤波与局域均值分解相结合能有效提取信号中的故障特征信息。(3)最后从描述信号复杂度的角度出发,采用基于LMD的多尺度近似熵方法,对轴承振动信号进行故障特征提取,该方法能有效区分不同的故障类型,比近似熵具有更强的抗干扰能力,获取更多的故障特征信息,仿真实验和实例分析表明,该方法可以应用到轴承故障特征提取中,判断轴承的运行状态。
[Abstract]:Rolling bearings are widely used in industrial production. As the core components of mechanical equipment, the running state of rolling bearings directly affects the reliability and stability of mechanical equipment.Therefore, the fault diagnosis of rolling bearings is of great significance to the operation and maintenance of mechanical equipment.The feature extraction of fault information is the key of bearing fault diagnosis. In this paper, local mean decomposition (LMD), morphological filtering and approximate entropy theory are used to diagnose rolling bearing faults.From the angle of signal filter denoising and signal sequence complexity, the vibration signal extraction method of rolling bearing is studied experimentally, which provides a theoretical basis for feature extraction of rolling bearing fault diagnosis.The main research work of this paper is as follows: (1) aiming at the non-stationary characteristic of rolling bearing vibration signal and the difficulty of extracting the actual fault characteristic signal, the local mean decomposition method is studied.This method can decompose the complex non-stationary signal into a series of amplitude modulation and frequency modulation functions, realize the separation of different modulation frequency components in the signal, and can effectively separate out the fault component.It can be applied to the decomposition of actual rolling bearing vibration signal. Aiming at the problem of serious noise interference of actual bearing vibration signal, the application of morphological filtering algorithm in vibration signal processing is studied.The filtered signal is decomposed into local mean to extract the fault feature information.Simulation results show that the combination of morphological filtering and local mean decomposition can effectively extract fault feature information from the signal. Finally, from the point of describing the complexity of the signal, the multi-scale approximate entropy method based on LMD is adopted.Fault feature extraction of bearing vibration signal shows that this method can distinguish different fault types effectively, has stronger anti-interference ability than approximate entropy, and obtains more fault feature information. The simulation experiment and example analysis show that the proposed method can effectively distinguish different fault types, and obtain more fault feature information than approximate entropy.This method can be applied to the bearing fault feature extraction to judge the bearing running state.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TH133.33

【参考文献】