基于振动信号的轴承状态监测和故障诊断方法研究

发布时间：2018-11-29 12:05

【摘要】：轴承是现代机械系统中相当重要的一类部件,其工作环境十分恶劣,使得轴承在运行时很容易发生故障。并且随着现代机械系统朝着大型化、精密化、高速化方向发展,轴承故障对整个系统的影响逐渐变大。因此,轴承故障诊断技术变得日趋重要。由于振动信号与轴承故障有直接相关性,并且基于振动信号的轴承故障诊断方法有诊断速度快、操作简单、可靠性强等优点。因此,该故障诊断方法以成为当前轴承故障诊断方法研究的热点之一。首先针对轴承故障振动信号频率成分复杂的特点,采用时域分析方法中的概率密度、有效值、峰值和峭度四种参数和频域中的功率谱方法分析该振动信号用以监测轴承的运行状态,判断轴承是否发生故障。之后针对时域分析方法无法诊断出轴承故障类型以及传统共振解调方法中的带通滤波器的通带较难选择的问题,采用经验模态分解和希尔伯特谱分析的共振解调的方法分析振动信号。先将信号分解成为多个本征模态函数之和,然后对于原振动信号相关性大的本征模态函数进行希尔伯特谱分析以求其包络谱,提取其故障特征频率。最后以美国凯斯西储大学轴承数据中心的故障数据为分析对象,对其进行状态监测和故障诊断。实验结果表明将时域参数分析中的峭度与其他三种时域参数分析结合可以很好地判断轴承是否发生故障,而共振解调方法能够诊断出轴承三种表面损伤类故障中的早期外圈和内圈故障。
[Abstract]:Bearing is a very important part in modern mechanical system, and its working environment is very bad. And with the development of modern mechanical system in the direction of large scale, precision and high speed, the influence of bearing fault on the whole system is gradually becoming greater. Therefore, bearing fault diagnosis technology has become increasingly important. Because of the direct correlation between vibration signal and bearing fault, the method of bearing fault diagnosis based on vibration signal has the advantages of fast diagnosis speed, simple operation, strong reliability and so on. Therefore, the fault diagnosis method has become one of the hotspots in the research of bearing fault diagnosis. Firstly, aiming at the complex frequency component of bearing fault vibration signal, the probability density and effective value in time domain analysis method are adopted. The four parameters of peak and kurtosis and the power spectrum method in frequency domain are used to analyze the vibration signal to monitor the running state of the bearing and to judge whether the bearing is in trouble or not. Then the time domain analysis method can not diagnose the bearing fault type and the traditional resonance demodulation method, which is difficult to select the passband of the band-pass filter. The resonance demodulation method based on empirical mode decomposition and Hilbert spectrum analysis is used to analyze the vibration signal. Firstly, the signal is decomposed into the sum of several intrinsic mode functions, and then the Hilbert spectrum analysis is carried out for the intrinsic mode function with high correlation of the original vibration signal to obtain its envelope spectrum, and the fault characteristic frequency is extracted. Finally, taking the fault data of bearing data center of case Western Reserve University as the analysis object, the condition monitoring and fault diagnosis are carried out. The experimental results show that the kurtosis in the time domain parameter analysis can be combined with the other three time domain parameter analysis to judge whether the bearing is in trouble or not. The resonance demodulation method can diagnose the early faults of outer ring and inner ring in three kinds of surface damage faults of bearing.
【学位授予单位】：中南大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH165.3

【相似文献】