滚动轴承故障诊断中信号校正和特征提取方法研究

发布时间：2018-04-26 22:27

本文选题：状态监测与故障诊断 + 滚动轴承　；参考：《中国科学技术大学》2017年硕士论文

【摘要】：机械设备的状态监测与故障诊断在保证产品质量、提高生产效率和避免安全事故等方面具有重大意义。滚动轴承作为各类旋转机械的重要部件,其健康状况直接影响了整个设备的性能,因此滚动轴承的故障诊断是机械设备故障诊断领域的重要课题。本文对非平稳信号的畸变校正以及强背景噪声下信号的特征提取方法进行了研究。首先,本文介绍了滚动轴承的基本结构、失效形式以及故障特征频率,分析了滚动轴承故障的发展历程、轴承振动的机理以及振动信号与轴承状态的关系。随后,进行了轴承振动信号采集实验和列车轴承动态声学信号采集实验。其中针对列车轴承动态声学信号采集实验,课题组设计了轴承静态声学实验和轴承动态声学实验来模拟列车的真实运动场景。其次,本文对列车道旁声学信号的多普勒畸变问题进行了研究。在麦克风阵列模型的基础上提出结合传播因子算法和时域插值重采样的多普勒畸变校正算法。首先利用传播因子算法对列车轴承的位置进行实时估计,随后根据轴承位置构造重采样时间序列,最后利用时域插值重采样方法对畸变信号进行矫正。利用模拟信号和实验信号对算法进行验证,可以看出该算法能有效地校正信号的多普勒畸变。最后,针对轴承振动信号在强背景噪声下的故障特征提取问题,本文提出了一种时变奇异值分解方法来提取轴承故障的频谱特征。该方法不需要像传统方法那样在分解后重构信号,而是直接从构造的时变奇异值序列中提取信号的故障特征频率。实验结果表明,该方法不仅在轴承的故障特征频率提取中效果显著,而且在轴承复合故障的分离中也有一定效果。
[Abstract]:The condition monitoring and fault diagnosis of mechanical equipment is of great significance in ensuring product quality, improving production efficiency and avoiding safety accidents. As an important part of all kinds of rotating machinery, the health condition of rolling bearing directly affects the performance of the whole equipment. Therefore, the fault diagnosis of rolling bearing is an important subject in the field of mechanical equipment fault diagnosis. In this paper, the distortion correction of nonstationary signal and the feature extraction method of the signal under strong background noise are studied. Firstly, this paper introduces the basic structure, failure form and fault characteristic frequency of rolling bearing, analyzes the development history of rolling bearing fault, the mechanism of bearing vibration and the relationship between vibration signal and bearing state. Then, the bearing vibration signal acquisition experiment and the train bearing dynamic acoustic signal acquisition experiment are carried out. For the dynamic acoustic signal acquisition experiment of train bearing, the static acoustic experiment and the dynamic acoustic experiment are designed to simulate the real moving scene of the train. Secondly, the Doppler distortion of the acoustic signal beside the train track is studied in this paper. Based on the microphone array model, a Doppler distortion correction algorithm combined with propagation factor algorithm and time domain interpolation resampling is proposed. Firstly, the transmission factor algorithm is used to estimate the position of the train bearing in real time, then the resampling time series is constructed according to the position of the bearing. Finally, the distortion signal is corrected by using the time domain interpolation resampling method. The algorithm is verified by analog signal and experimental signal. It can be seen that the algorithm can effectively correct the Doppler distortion of the signal. Finally, a time-varying singular value decomposition method is proposed to extract the fault spectrum feature of bearing vibration signal under strong background noise. This method does not need to reconstruct the signal after decomposing as the traditional method, but directly extracts the fault characteristic frequency of the signal from the constructed time-varying singular value sequence. The experimental results show that this method is effective not only in the extraction of bearing fault feature frequency, but also in the separation of bearing composite fault.
【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】