【摘要】 为提取机械设备故障诊断中的某些微弱信号,给出一种新的提取强背景噪声中微弱周期信号的方法.首先对原始信号进行经验模态分解,得到理论意义上的固有模态函数;然后对各分解层做自相关分析,依据自相关图像,对可能含有周期成分的分解层进行频谱分析;最后可提取微弱周期信号的频率及幅值信息.通过仿真分析,证实该方法能够有效提取淹没于背景噪声中的微弱周期信号.更多还原

【Abstract】 A new approach to extract weak periodic signal from strong noise is proposed to pick up some weak signals in the fault diagnosis of mechanical equipment. Original signal is firstly decomposed by empirical mode decomposition method, and then autocorrelation analysis is made respectively for consequent intrinsic mode functions. Some intrinsic mode functions containing periodic components are selected according to corresponding autocorrelogram and further analyzed by frequency spectrum, thus the weak periodic signal can be extracted. Simulation results show that the approach can effectively extract the weak periodic signals hidden in the strong background noise.更多还原