通过仿真实验将小波变换和经验模态分解(EMD)方法分解信号的能力进行了比较,并将这种滤波特性应用于旋转机械的故障诊断中,结合包络谱分析,比较了两者对于滚动轴承内圈故障的诊断效果.仿真及轴承实验结果表明EMD方法在滤波的自适应性、分解结果的准确性以及诊断效果等方面均具有优势,更重要的是它分离出的主要分量物理意义明确,反映了信号的真实内涵. Through the simulation experiments, the ability of wavelet transform and empirical mode decomposition (EMD) method to decompose signal is compared. And this filtering characteristic is applied to the fault diagnosis of rotating machinery. Combined with the envelope spectrum analysis, The results of simulation and bearing experiments show that the EMD method has advantages in terms of adaptive filtering, the accuracy of the decomposition results and the diagnostic effect, and more importantly, the main component of the EMD method has a clear physical meaning, Reflects the true meaning of the signal.