基于循环平稳的电机轴承故障特征分析

发布时间：2018-11-09 14:33

【摘要】：三相感应电机结构简单、运行可靠且高效,被广泛地应用于生产生活的各个领域。滚动轴承作为电机的重要组成部分很容易发生故障甚至会造成严重的后果。因此,对轴承进行定期的故障检测及维护,及早的发现故障并采取措施是至关重要的。本文针对轴承实际故障时损伤区域的大小、轴承动力学结构以及滚珠进出坑时的承载力变化等情况,提出了轴承故障时的双脉冲转矩波动模型,并基于该模型推导了定子电流中特征频率的表达式。通过对双脉冲模型下的定子电流信号分别进行循环自相关函数和循环谱密度函数分析,相较于传统的轴承故障特征提取与识别方法,发现循环平稳理论在基频及谐波的降噪、轴承故障特征频率的识别与提取方面有明显的优越性。本文研究了轴承故障特征频率的幅值与故障损伤宽度的关系,利用循环自相关函数分析了不同故障损伤宽度下,轴承故障特征频率第一至第三边频的幅值变化规律。然后对定子电流信号进行循环谱密度分析,可以看出其在循环频率域与谱频率域具有明显的谱相关性,利用这种谱相关性可以从不同角度识别出故障特征,不仅仅能反映出循环自相关的信息,同时也会在谱频率域中显示出更多的故障特征信息,为轴承故障的识别提供了更多的判断依据。在实验室环境下利用电机轴承故障实验平台采集了电机不同故障宽度下的定子电流信号,并对这些电流信号进行循环自相关函数和循环谱密度函数分析。将实际分析结果与仿真结果进行对比,验证了本文理论和方法的正确性,也显示了循环平稳理论的信号处理方法在识别和提取轴承故障特征频率分量方面的优越性。
[Abstract]:Three-phase induction motor with simple structure, reliable operation and high efficiency is widely used in various fields of production and life. As an important part of motor, rolling bearings are prone to failure and even serious consequences. Therefore, it is very important to detect and maintain the bearing regularly, to detect the fault as early as possible and to take measures. In this paper, a double pulse torque ripple model is proposed for bearing failure, such as the size of damage zone, the dynamic structure of bearing, and the change of bearing capacity when ball is entering or leaving the pit. Based on the model, the expression of characteristic frequency in stator current is derived. Through the analysis of the stator current signal under the double pulse model by the cyclic autocorrelation function and the cyclic spectral density function, compared with the traditional method of bearing fault feature extraction and identification, it is found that the cyclic stationary theory reduces the noise of the fundamental frequency and harmonics. Bearing fault feature frequency recognition and extraction has obvious advantages. In this paper, the relationship between the amplitude of bearing fault characteristic frequency and the fault damage width is studied. By using cyclic autocorrelation function, the variation of amplitude of bearing fault characteristic frequency from the first to the third edge frequency under different fault damage widths is analyzed. Then, by analyzing the cyclic spectral density of stator current signal, it can be seen that it has obvious spectral correlation between cyclic frequency domain and spectral frequency domain, and the fault characteristics can be identified from different angles by using this spectral correlation. It can not only reflect the cyclic autocorrelation information, but also show more fault characteristic information in the spectrum frequency domain, which provides more judgment basis for bearing fault identification. The stator current signals under different fault widths of motor are collected by using the motor bearing fault test platform in the laboratory, and the cyclic autocorrelation function and cyclic spectral density function are used to analyze these current signals. The comparison between the actual analysis results and the simulation results verifies the correctness of the theory and method in this paper, and also shows the superiority of the signal processing method of the cyclic stationary theory in identifying and extracting the frequency components of bearing fault characteristics.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM307

【参考文献】