基于多尺度随机共振谱的滚动轴承故障诊断方法研究

发布时间：2018-01-20 17:43

本文关键词： 轴承故障诊断随机共振时频分析多尺度噪声调节算法软件实现　出处：《中国科学技术大学》2017年硕士论文　论文类型：学位论文

【摘要】：旋转设备是工业机械中最重要的部件之一,而轴承则是一种典型的旋转设备。如果机械系统中的轴承有故障,不仅会影响系统的正常运行,也有可能导致一些意想不到的危险后果,所以需要及时发现轴承的早期缺陷。一般通过分析振动信号来进行轴承的状态检测和故障诊断,但是受到工作环境的噪声及轴承与其他机械零件耦合而产生的噪声的影响,故障诊断的难度很大。因此,亟需找到一种有效的微弱信号检测的方法。在微弱信号检测中,随机共振是一种利用噪声来增强周期性信号的有效方法。考虑到有故障的轴承振动信号的非平稳特性,本文研究了时频分布中的随机共振。本文提出了一种称为多尺度随机共振谱的新方法,以提高检测初始缺陷信号的有效性。该方法的理论依据在于(1)故障导致的瞬变主要位于时频分布中的特定频带,因此只有该频段的噪声才能激活随机共振效应;(2)时频分布上的每一个频率上对应于在特定频率上调制的包络,因此在时频分布中存在用于每个频率标度的调制系统。本文提出的新方法将时频分布的每个尺度视为特定频率的调制系统,由于有用信息仅包含在测量信号的特定尺度中,不同调制系统中的噪声在用随机共振技术增强缺陷信息方面将具有不同的有效性。在对实验数据的分析中,相比起各种经典的随机共振方法,本文所提出的方法在识别滚动轴承的故障频率方面具有更多的优势,同时也显示了在诊断混合故障的振动信号中的潜力。此外,本文还探索了两种机械故障诊断算法的软件实现,其一是在智能手机中应用故障诊断算法,其二是将算法部署到服务器上并通过网络浏览器将诊断结果可视化,这两种方法分别实现了在线与离线的故障诊断。
[Abstract]:Rotating equipment is one of the most important components in industrial machinery, while the bearing is a kind of typical rotating equipment. If the bearings in the mechanical system fault, will not only affect the normal operation of the system, may also lead to some unexpected and dangerous consequences, so they need to find out the early defect of bearing. By analysis of vibration signal to carry out detection and fault diagnosis of the bearing condition, but by the impact of noise and noise of bearing working environment and other mechanical parts generated by a combination of the fault diagnosis is very difficult. Therefore, it is urgent to find an effective method of weak signal detection. In weak signal detection, stochastic resonance is a kind of effective enhancement the method of periodic signal with noise. Considering the non-stationary characteristics of vibration signals of bearing fault, this paper studies the time-frequency distribution of stochastic resonance is proposed in this paper. A new method called random resonance spectra of multi-scale, in order to improve the effectiveness of the detection of initial defect signal. The theoretical basis of this method is that (1) caused by the fault transient mainly in the specific frequency band in frequency distribution, so the band noise can only activate the stochastic resonance effect; (2) when the envelope of each frequency distribution on the frequency corresponding to the modulation at a particular frequency, so the time-frequency distribution in frequency modulation system for each scale. A new method is put forward in this paper at each scale frequency distribution as a particular frequency modulation system, the useful information contained in a specific scale the measurement signal in the noise of different modulation system in the enhanced defect information will have different effective use of stochastic resonance technology. In the analysis of the experimental data, compared with the classical method of stochastic resonance, the Has the advantage of fault frequency method of rolling bearing in the recognition, but also shows the vibration signal in fault diagnosis of mixed potential. In addition, this paper also explores the two kinds of mechanical fault diagnosis algorithm of the software, it is should be used in the fault diagnosis algorithm of intelligent mobile phone, the second is the algorithm deployed to the server through a web browser and the diagnostic results visualization, these two methods are used to implement the fault diagnosis online and offline.

【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TH133.33

【参考文献】