基于HHT和SVM的滚动轴承故障振动信号的诊断研究
发布时间：2019-04-21 20:33
【摘要】：滚动轴承作为机械设备中一个重要的组成部分,对其进行状态检测和故障诊断具有很强的现实意义。本文利用希尔伯特-黄变换法(HHT)对滚动轴承故障信号进行能量特征值提取,进而利用支持向量机(SVM)的方法对滚动轴承故障状态进行识别。 滚动轴承故障诊断主要包括诊断信息的获取,故障特征值的提取和模式识别三个部分。其中故障特征的提取和状态识别是滚动轴承故障诊断的关键。当滚动轴承发生故障时,其振动信号往往表现为非平稳性,本文提出的希尔伯特-黄变换法中的EMD分解法是基于信号的局部时间特征尺度,具有很强的自适应性,可以将信号分解为有限个本征模函数(IMF)之和,每个IMF分量分别包括了不同时间特征尺度大小的成分,其尺度依次由小到大,因此,每个IMF分量包含了从高到低不同频率段信号成分。本文将EMD方法引入滚动轴承故障诊断,选取故障信息明显的IMF分量,提取出其能量特征向量,实现了滚动轴承故障的初步诊断。 在对滚动轴承进行模式识别上本文采用了支持向量机方法,因它具有对经验的依赖小,能够获得全局最优解以及良好的泛化性能等特点,已被广泛应用于模式识别中。本文将提取到的IMF分量的能量特征向量作为支持向量机的输入从而进行分类应用于滚动轴承故障诊断识别中,实现了对滚动轴承故障状态准确的诊断识别。
[Abstract]:As an important part of mechanical equipment, rolling bearing has strong practical significance in condition detection and fault diagnosis. In this paper, the Hilbert-Huang transform (HHT) method is used to extract the energy eigenvalues of the rolling bearing fault signal, and then the support vector machine (SVM) is used to identify the fault state of the rolling bearing. The fault diagnosis of rolling bearing consists of three parts: the acquisition of diagnosis information, the extraction of fault characteristic value and the pattern recognition. Fault feature extraction and state recognition are the key points of rolling bearing fault diagnosis. When the rolling bearing fails, the vibration signal is usually non-stationary. The EMD decomposition method proposed in this paper is based on the local time characteristic scale of the signal, and has strong self-adaptability. The signal can be decomposed into the sum of a finite number of eigenmode functions (IMF). Each IMF component includes components of different time characteristic scales, and their scales are small to large in turn, so each IMF component includes components of different time characteristic scales, so the scale of each IMF component is from small to large. Each IMF component contains signal components from high to low frequencies. In this paper, the EMD method is introduced into the fault diagnosis of rolling bearing, and the IMF component with obvious fault information is selected. The energy characteristic vector is extracted, and the primary diagnosis of rolling bearing fault is realized. In this paper, the support vector machine (SVM) method is used in the pattern recognition of rolling bearings. It has been widely used in pattern recognition because of its small dependence on experience, the ability to obtain the global optimal solution and the good generalization performance. In this paper, the energy eigenvector of the extracted IMF component is used as the input of the support vector machine and applied to the fault diagnosis and recognition of the rolling bearing, and the accurate diagnosis and recognition of the fault state of the rolling bearing is realized.
【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;TP18
本文编号：2462532
[Abstract]:As an important part of mechanical equipment, rolling bearing has strong practical significance in condition detection and fault diagnosis. In this paper, the Hilbert-Huang transform (HHT) method is used to extract the energy eigenvalues of the rolling bearing fault signal, and then the support vector machine (SVM) is used to identify the fault state of the rolling bearing. The fault diagnosis of rolling bearing consists of three parts: the acquisition of diagnosis information, the extraction of fault characteristic value and the pattern recognition. Fault feature extraction and state recognition are the key points of rolling bearing fault diagnosis. When the rolling bearing fails, the vibration signal is usually non-stationary. The EMD decomposition method proposed in this paper is based on the local time characteristic scale of the signal, and has strong self-adaptability. The signal can be decomposed into the sum of a finite number of eigenmode functions (IMF). Each IMF component includes components of different time characteristic scales, and their scales are small to large in turn, so each IMF component includes components of different time characteristic scales, so the scale of each IMF component is from small to large. Each IMF component contains signal components from high to low frequencies. In this paper, the EMD method is introduced into the fault diagnosis of rolling bearing, and the IMF component with obvious fault information is selected. The energy characteristic vector is extracted, and the primary diagnosis of rolling bearing fault is realized. In this paper, the support vector machine (SVM) method is used in the pattern recognition of rolling bearings. It has been widely used in pattern recognition because of its small dependence on experience, the ability to obtain the global optimal solution and the good generalization performance. In this paper, the energy eigenvector of the extracted IMF component is used as the input of the support vector machine and applied to the fault diagnosis and recognition of the rolling bearing, and the accurate diagnosis and recognition of the fault state of the rolling bearing is realized.
【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;TP18
【引证文献】
相关博士学位论文 前1条
1 胡劲松;面向旋转机械故障诊断的经验模态分解时频分析方法及实验研究[D];浙江大学;2003年
相关硕士学位论文 前3条
1 范超;旋转机械振动故障信号微弱特征提取方法研究[D];东北石油大学;2013年
2 莫嘉林;基于代价敏感布雷格曼散度的旋转机械轴承故障诊断研究[D];长沙理工大学;2013年
3 周亮;矿用绞车滚动轴承故障诊断系统设计[D];重庆大学;2014年
，本文编号：2462532
本文链接：https://www.wllwen.com/kejilunwen/jixiegongcheng/2462532.html
