中文摘要：

针对在强噪声背景下轴承振动信号的非线性、非平稳性以及信号出现的复杂调制现象,提出了一种基于小波包熵与EEMD（总体模态分解）相结合的能量算子解调故障诊断方法.根据信号的小波包熵值对信号小波包降噪,采用EEMD方法选取相关度最大的IMF（本征模态分量）以避免选取信号的盲目性,对该分量进行能量算子解调,实现提取该分量下的故障信号的幅值和频率信息.对机械故障振动信号进行实验分析表明,相对于EMD分析存在的模态混叠现象以及普通Hilbert解调法的运算精度满足不了诊断需求的情况,该方法能够有效解调出故障频率信息,以实现对故障类别的推断.

英文摘要：

Aimed at the nonlinearity, unsteadiness, and complex modulation of bearing vibration signal in strong noise background, a fault diagnosis method is presented with energy operator demodulation based on wavelet package entropy and EEMD （ensemble empirical mode decomposition） Combined. In this meth- od, the wavelet package entropy value of the signal is denoised first and then by using EEMD method, the component of IMF （intrinsic mode function） with maximum correlativity is selected in order to avoid the blindness of signal selecting. Further, the IMF is demodulated with energy operator demoduiati0n so as to realize the extraction of the information of the amplitudeand frequency of the fault signal. Experimental a- nalysis of mechanical fault vibration signal shows that, as contrasted to the mixed-overlapping of mode in EMD analysis and the operational precision with ordinary Hilbert demodulation method which cannot satis- fy the requirement of diagnosis, this method can effectively demodulate the failure frequency information, so that the inference of the fault category will be realized.