【摘要】 经验模态分解(EMD)是由Huang等人提出的一种全新的针对非线性非平稳信号处理的算法.通过EMD,可以把一个信号分解为若干个固有模态函数(IMF),再将这些IMF进行希尔波特变换,从而得到具有真正意义的瞬时频率,因此解决了传统信号处理方法的不足之处。与此同时,EMD算法是一个全新的算法,本身也存在不足,如端点问题。文章在现有的解决方法的基础上,提出了用正交多项式拟合的方法来解决EMD的端点问题,并通过和已有算法的比较来证明这种方法的有效性。更多还原

【Abstract】 The Empirical Mode Decomposition( EMD) has been developed by Huang etc, which is a new method foranalyzing nonlinear and non- stationary signal.A signal can be decomposed into some Intrinsic Mode Function(IMF),which is processed by Hilbert transform for obtaining meaningful instantaneous frequency.Therefore, this new method hasresolved deficiencies belonging to traditional methods for processing signal.At the same time, this new method has somedeficiencies due to have been developed lately.One of the deficiencies is end issue.In our paper, based on existentmethods, we put forward Orthogonal Polynomial Fitting Algorithm to deal with this issue.We have proved our mean isavailable via comparing it with other means.更多还原