基于小波包能量谱齿轮振动信号的分析与故障诊断

Vibration Signal Analysis and Fault Diagnosis of Gear based on Wavelet Packet Energy Spectrum

小波包是继小波分析之后提出的一种新型的多尺度分析方法,解决了小波分析在高频部分分辨率差的缺点,体现了比小波分析更好的处理效果。测试了齿轮传动系统在几种不同故障类型下的振动信号,利用小波包变换的分解和重构算法,有效地提取出齿轮故障特征信号,得到试验结果。通过比较时域分析、频域分析和小波包分析对齿轮振动信号进行的特征提取,从能量分布的角度出发,阐述了在齿轮故障诊断中,利用小波包分析方法将故障信号进行进一步量化、分解,从而进行判断的方法。试验表明了基于小波包能量谱分析在齿轮故障信号提取与诊断中的优势。

The wavelet packet transform is a new multi-scale analysis method developed in the past decade.It overcomes the shortcomings of poor resolution in high-frequency part inherent in wavelet analysis.Vibration fault feature of gear is analyzed.By using the decomposition and reconstruction algorithm of wavelet transform,the fault feature signal of gear are extracted availably and the result s of experimentation are obtained.The characteristics of time domain analysis,frequency spectrum analysis and wavelet analysis are compared,and the predominance of wavelet energy spectrum analysis in extraction and diagnosis of fault signal is proved.