变分模态分解与神经网络结合的轴承故障诊断
发布时间：2019-06-04 01:18
【摘要】：故障信号特征提取的准确性是保证故障智能诊断识别率的关键因素。低信噪比情况下,故障诊断效果下降。变分模态分解方法(VMD)在信号分解精度和抗噪方面具有明显优势。在分析VMD抗噪性能的基础上,提出以VMD分解的各模态能量作为智能诊断特征信息,并与小波包的特征信息进行对比研究。将滚动轴承两种故障特征信息通过BP神经网络识别,用不同信噪比的加噪故障信号进行测试,结果表明,在低信噪比情况下基于VMD模态能量的故障特征更具有可识别性。
[Abstract]:The accuracy of fault signal feature extraction is the key factor to ensure the recognition rate of intelligent fault diagnosis. Under the condition of low signal-to-noise ratio (SNR), the effect of fault diagnosis is decreased. (VMD) has obvious advantages in signal decomposition accuracy and anti-noise. On the basis of analyzing the anti-noise performance of VMD, the modal energy of VMD decomposition is proposed as the characteristic information of intelligent diagnosis, and the characteristic information of wavelet packet is compared with that of wavelet packet. Two kinds of fault feature information of rolling bearing are identified by BP neural network and tested with different signal-to-noise ratio (SNR). The results show that the fault feature based on VMD modal energy is more identifiable in the case of low SNR.
【作者单位】： 上海开放大学信息与工程学院;上海大学机电工程与自动化学院;
【基金】：国家自然科学基金资助项目(51575331)
【分类号】：TH133.33
本文编号：2492369
[Abstract]:The accuracy of fault signal feature extraction is the key factor to ensure the recognition rate of intelligent fault diagnosis. Under the condition of low signal-to-noise ratio (SNR), the effect of fault diagnosis is decreased. (VMD) has obvious advantages in signal decomposition accuracy and anti-noise. On the basis of analyzing the anti-noise performance of VMD, the modal energy of VMD decomposition is proposed as the characteristic information of intelligent diagnosis, and the characteristic information of wavelet packet is compared with that of wavelet packet. Two kinds of fault feature information of rolling bearing are identified by BP neural network and tested with different signal-to-noise ratio (SNR). The results show that the fault feature based on VMD modal energy is more identifiable in the case of low SNR.
【作者单位】： 上海开放大学信息与工程学院;上海大学机电工程与自动化学院;
【基金】：国家自然科学基金资助项目(51575331)
【分类号】：TH133.33
【相似文献】
相关博士学位论文 前1条
1 吕中亮;基于变分模态分解与优化多核支持向量机的旋转机械早期故障诊断方法研究[D];重庆大学;2016年
相关硕士学位论文 前2条
1 王敬;基于离散模态分解法的表面建模与装配仿真研究[D];浙江大学;2017年
2 王振威;基于变分模态分解的故障诊断方法研究[D];燕山大学;2015年
，本文编号：2492369
本文链接：https://www.wllwen.com/jixiegongchenglunwen/2492369.html
