机电设备振动信号故障诊断算法研究

发布时间：2018-06-05 02:20

本文选题：特征提取 + 小波包分解　；参考：《计算机仿真》2017年05期

【摘要】：研究机电设备振动信号故障诊断问题,由于运行在复杂工况下的故障信号很容易淹没在噪声中,而传统的特征提取方法不能较好地对特征信号准确提取,降低了故障诊断的准确性。针对上述问题,提出了基于频带方差的小波包分解与参数优化的支持向量机的故障诊断算法。首先利用振动信号的小波包分解系数计算各子频带的方差,并作为特征向量。然后将特征向量作为支持向量机的输入优化其参数,对故障进行"多对多""的分类。仿真结果表明,提出的算法提高了机电设备振动信号故障诊断的准确率。
[Abstract]:The problem of fault diagnosis of vibration signal of electromechanical equipment is studied. Because the fault signal running under complex working conditions is easily submerged in noise, the traditional feature extraction method can not extract the characteristic signal accurately. The accuracy of fault diagnosis is reduced. Aiming at the above problems, a fault diagnosis algorithm based on wavelet packet decomposition and parameter optimization for support vector machines based on band variance is proposed. First, the wavelet packet decomposition coefficient of vibration signal is used to calculate the variance of each subband, and it is used as the eigenvector. Then the feature vector is used as the input of the support vector machine to optimize its parameters, and the fault is classified as "many-to-many". Simulation results show that the proposed algorithm improves the accuracy of vibration signal fault diagnosis of electromechanical equipment.
【作者单位】：北京信息科技大学信息与通信工程学院;
【基金】：北京市科技创新提升计划项目(PXM2016_014224_000021)
【分类号】：TH17

【相似文献】