基于遗传神经网络的滚动轴承故障诊断方法的研究
发布时间：2019-01-21 17:20
【摘要】：在机械设备故障中因为滚动轴承受损而发生故障的概率非常大,滚动轴承的工作状况将直接影响到整个机械设备的运转。滚动轴承故障的诊断方法有很多种,随着科学技术的发展与进步,对滚动轴承检测的实时性和诊断结果、维修方案的准确性等提出了新的要求,这对于滚动轴承故障诊断的研究具有非常重要的意义。 为了能够更好的对滚动轴承故障进行诊断,本文分析了当前应用于滚动轴承故障诊断的主要方法,应用当前使用较多的小波分析理论,对轴承的振动信号进行阈值降噪,并提出了较为合理的新阈值函数,对新阈值函数进行了理论分析,实验结果表明,新阈值函数对含噪信号的降噪效果,比传统阈值函数的降噪效果更加优越。同时,采用小波包的方法提取轴承故障特征信息,将BP神经网络智能化诊断引入诊断系统中,以小波包提取的特征信息作为BP神经网络训练样本和预测样本。因BP神经网络自身存在易于陷入局部极小值和收敛速度慢等缺陷,采用遗传算法对BP神经网络的初始权值和阈值进行优化。优化的BP神经网络可以较好地克服BP网络的缺陷,在滚动轴承故障训练和诊断时,可以找到全局最优值。采用LabVIEW友好界面的开发功能和MATLAB强大的数值分析和数据处理的功能,进行滚动轴承故障诊断系统的研发,充分利用LabVIEW自带的MATLAB script节点,将两种软件的优点结合到一起,实现了滚动轴承的智能化诊断,这也使得该系统对故障的诊断速度和准确度得到较大的提高。
[Abstract]:The fault probability of the rolling bearing is very large in the mechanical equipment fault. The working condition of the rolling bearing will directly affect the operation of the whole machinery and equipment. There are many kinds of fault diagnosis methods for rolling bearings. With the development and progress of science and technology, new requirements are put forward for the real-time detection and diagnostic results of rolling bearings, the accuracy of maintenance schemes, etc. This is of great significance to the research of rolling bearing fault diagnosis. In order to diagnose the fault of rolling bearing better, this paper analyzes the main methods of fault diagnosis of rolling bearing at present, and applies more wavelet analysis theory to reduce the noise of the vibration signal of bearing. A reasonable new threshold function is put forward and the theoretical analysis of the new threshold function is made. The experimental results show that the new threshold function is more effective than the traditional threshold function in reducing the noise of the noisy signal. At the same time, the fault feature information of bearing is extracted by wavelet packet method, and the intelligent diagnosis of BP neural network is introduced into the diagnosis system. The feature information extracted by wavelet packet is used as the training sample and prediction sample of BP neural network. Because the BP neural network is easy to fall into the local minimum and the convergence speed is slow, the genetic algorithm is used to optimize the initial weights and thresholds of the BP neural network. The optimized BP neural network can overcome the defects of BP neural network and find the global optimal value in the fault training and diagnosis of rolling bearing. By adopting the development function of LabVIEW friendly interface and the powerful function of numerical analysis and data processing of MATLAB, the research and development of rolling bearing fault diagnosis system are carried out, and the advantages of the two kinds of software are combined by making full use of the MATLAB script node of LabVIEW. The intelligent diagnosis of rolling bearing is realized, which improves the speed and accuracy of fault diagnosis.
【学位授予单位】：山东理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP183;TH165.3
本文编号：2412858
[Abstract]:The fault probability of the rolling bearing is very large in the mechanical equipment fault. The working condition of the rolling bearing will directly affect the operation of the whole machinery and equipment. There are many kinds of fault diagnosis methods for rolling bearings. With the development and progress of science and technology, new requirements are put forward for the real-time detection and diagnostic results of rolling bearings, the accuracy of maintenance schemes, etc. This is of great significance to the research of rolling bearing fault diagnosis. In order to diagnose the fault of rolling bearing better, this paper analyzes the main methods of fault diagnosis of rolling bearing at present, and applies more wavelet analysis theory to reduce the noise of the vibration signal of bearing. A reasonable new threshold function is put forward and the theoretical analysis of the new threshold function is made. The experimental results show that the new threshold function is more effective than the traditional threshold function in reducing the noise of the noisy signal. At the same time, the fault feature information of bearing is extracted by wavelet packet method, and the intelligent diagnosis of BP neural network is introduced into the diagnosis system. The feature information extracted by wavelet packet is used as the training sample and prediction sample of BP neural network. Because the BP neural network is easy to fall into the local minimum and the convergence speed is slow, the genetic algorithm is used to optimize the initial weights and thresholds of the BP neural network. The optimized BP neural network can overcome the defects of BP neural network and find the global optimal value in the fault training and diagnosis of rolling bearing. By adopting the development function of LabVIEW friendly interface and the powerful function of numerical analysis and data processing of MATLAB, the research and development of rolling bearing fault diagnosis system are carried out, and the advantages of the two kinds of software are combined by making full use of the MATLAB script node of LabVIEW. The intelligent diagnosis of rolling bearing is realized, which improves the speed and accuracy of fault diagnosis.
【学位授予单位】：山东理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP183;TH165.3
【参考文献】
相关期刊论文 前3条
1 谢培甫;;基于小波分析和BP神经网络的滚动轴承的故障诊断[J];农业装备与车辆工程;2006年02期
2 刘春光;谭继文;张驰;;基于小波分析的滚动轴承的故障特征提取技术[J];机械工程与自动化;2010年02期
3 付军,胡晓依,陈亮,唐小兵;BP神经网络方法在机车轴承故障诊断中的应用[J];铁道机车车辆;2002年02期
，本文编号：2412858
本文链接：https://www.wllwen.com/kejilunwen/jixiegongcheng/2412858.html
