基于经验模态分解的混沌噪声背景下弱信号检测与信号提取

基于经验模态分解方法,研究了在强混沌噪声背景下进行弱信号的检测与信号提取。对仿真信号的研究表明:用该方法可以直接提取出微弱的偶然性和周期性冲击时域信号,对弱谐波信号可能不能直接提取,但可以直接提取出其频率特征,这些弱冲击信号和弱谐波信号完全淹没在强的混沌噪声背景信号中,无论从时域上还是频域上基本上都看不出来。对齿轮箱的实际信号的研究也表明:尽管某些故障信号有时极其微弱,EMD方法也能有效地实现这些非线性非平稳信号的分离和提取,从而为机械设备故障诊断提供直观的有效的参考。

Weak Signal Detection and Extraction Against the Background of Chaotic Noise Based on Empirical Mode Decomposition

A new approach to detecting and extracting weak fault signals on the basis of empirical mode decomposition(EMD) against the background of chaotic noise is presented.Studies of simulated signals show that weak occasional and periodic impulse signals can be extracted directly by using the approach,but weak harmonic signals may not be extracted directly,but their frequency characteristics can be extracted,while these signals are totally flooded by background signals of strong chaotic noise,and are hardly visible in time or frequency domains.Studies of gearbox vertual signals also demonstrate that some fault signals,though extremely weak,may be effectively detected and extracted.Therefore,the approach offers an intuitive and effective reference to machinery fault diagnosis.