基于声发射的轴承滚动接触疲劳量化诊断技术研究

发布时间：2018-04-09 19:17

本文选题：声发射　切入点：轴承滚动接触疲劳　出处：《燕山大学》2016年硕士论文

【摘要】：在常规运行条件下,轴承发生滚动接触疲劳失效是最常见的轴承故障之一。轴承从出现接触疲劳现象到失效是一个动态发展过程,因此,从安全性和经济性两方面综合考虑,对轴承滚动接触疲劳状态进行实时监测并及时作出诊断具有重要意义。本文结合国内外基于声发射技术的轴承滚动接触疲劳诊断的研究现状,提出了一种用于量化诊断轴承滚动接触疲劳状态的监测系统方案并进行了相关试验研究。本文主要从以下几个方面开展研究工作:(1)研究了轴承滚动接触疲劳声发射信号的相关处理方法。主要包括:声发射信号的预处理方法,主要应用小波包分解和经验模态分解法对声发射信号进行滤波消噪;声发射信号的特征提取方法,研究应用小波包分析分别提取声发射信号的能量特征与信息熵特征;声发射信号的特征选择,应用主成分分析法对提取的声发射信号的特征进行压缩降维;声发射信号特征的模式识别方法,主要应用支持向量机和BP神经网络对经选择后的特征进行模式识别,判断轴承的滚动接触疲劳状态。(2)设计了基于信息融合技术的轴承滚动接触疲劳量化诊断方案。应用“特征串行”法这一中间层次的特征级信息融合方法,将两通道声发射传感器获取的声发射信号融合在一起进行轴承滚动接触疲劳状态的判别。(3)设计并建立了基于声发射技术的轴承滚动接触疲劳量化诊断系统。结合实验室现有设备和技术条件搭建了量化诊断系统的硬件部分并对诊断系统的软件部分进行了功能、结构和必要的算法设计。(4)进行试验以验证该量化诊断系统的准确性和可靠性。通过对试验过程中获取到的声发射信号数据进行分析和处理得出诊断结果,将诊断结果和轴承试件样本进行对比以验证该量化诊断系统的准确性。
[Abstract]:Rolling contact fatigue failure is one of the most common bearing failures under normal operating conditions.Bearing contact fatigue phenomenon to failure is a dynamic development process, so it is important to monitor and diagnose rolling contact fatigue state in real time from safety and economy aspects.In this paper, according to the research status of rolling contact fatigue diagnosis based on acoustic emission technology at home and abroad, a monitoring system for quantitative diagnosis of rolling contact fatigue state of bearings is proposed and related experimental research is carried out.In this paper, the related processing method of bearing rolling contact fatigue acoustic emission signal is studied from the following aspects: 1.It mainly includes: preprocessing method of acoustic emission signal, mainly using wavelet packet decomposition and empirical mode decomposition to filter and eliminate noise of acoustic emission signal, feature extraction method of acoustic emission signal,Wavelet packet analysis is used to extract the energy characteristics and information entropy characteristics of acoustic emission signals, the feature selection of acoustic emission signals and the principal component analysis (PCA) are used to compress and reduce the dimension of the extracted acoustic emission signals.The pattern recognition method of acoustic emission signal is mainly based on support vector machine (SVM) and BP neural network (BP neural network).The quantitative diagnosis scheme of bearing rolling contact fatigue based on information fusion technology is designed.Using the "feature serial" method, which is the middle level feature level information fusion method,The acoustic emission signals obtained from two acoustic emission sensors are fused together to distinguish the rolling contact fatigue state of bearings. A quantitative diagnosis system for rolling contact fatigue of bearings based on acoustic emission technology is designed and established.Combined with the existing equipment and technical conditions of the laboratory, the hardware part of the quantitative diagnosis system is built and the function of the software part of the diagnosis system is carried out.Structure and necessary algorithm design. 4) to verify the accuracy and reliability of the quantitative diagnostic system.The diagnostic results are obtained by analyzing and processing the acoustic emission signal data obtained during the test. The accuracy of the quantitative diagnosis system is verified by comparing the diagnostic results with the bearing specimen samples.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TH133.3

【相似文献】