该方法的关键是经验模式分解，它能使复杂信号分解为有限个本征模函数（Intrinsic Mode Function，简称IMF），所分解出来的各IMF分量包含了原信号的不同时间尺度的局部特征信号。经验模态分解法能使非平稳数据进行平稳化处理，然后进行希尔伯特变换获得时频谱图，得到有物理意义的频率。与短时傅立叶变换、小波分解等方法相比，这种方法是直观的、直接的、后验的和自适应的，因为基函数是由数据本身所分解得到。由于分解是基于信号序列时间尺度的局部特性，因此具有自适应性。(The key of this method is empirical mode decomposition, which decomposes the complex signal into a finite number of intrinsic mode functions (IMFs). Each IMF component is decomposed into local characteristic signals of different time scales of the original signal . Empirical mode decomposition method can stabilize the non-stationary data, and then perform Hilbert transform to obtain the time-frequency spectrum to obtain the physical frequency. Compared with short-time Fourier transform, wavelet decomposition and other methods, this method is intuitive, direct, posterior and adaptive, because the basis function is derived from the data itself. Since decomposition is based on the local characteristics of the time series of the signal sequence, it is self-adaptive.)