摘要
针对海上石油平台透平发电机振动信号的非平稳特性以及获得故障诊断样本数据较困难的特点,该文提出了一种基于小波包分析和SVM的故障诊断方法。首先将采样信号通过小波包分析去噪,即通过Mallat塔式算法对信号进行小波分解再重构,重构后的故障诊断子频带信号通过EMD算法提取故障诊断特征向量,并以此训练SVM。试验结果表明基于小波包分析和SVM的方法具有较高的精度和较高的诊断效率。
According to the non-stationary characteristics of offshore oil platform and turbine generator vibration signal characteristics is difficult to fault diagnosis of the sample data. This paper presents a fault diagnosis method based on wavelet packets analysis and SVM. Firstly,the sampling signal by wavelet analysis denouncing,the wavelet decomposition of the signal is reconstructed by Mallat tower algorithm. The fault diagnosis of the sub-band signal reconstruction by the EMD algorithm to extract the characteristic vector of fault diagnosis,and to train the SVM. The results show that the method based on wavelet packet analysis and SVM has higher precision and higher diagnosis efficiency.
出处
《自动化与仪表》
2018年第2期54-58,共5页
Automation & Instrumentation
