基于小波去噪和EMD的信号瞬时参数提取

为了消除随机噪声对经验模式分解（EMD）质量的影响，提出利用小波去噪作为EMD的预处理，并结合希耳伯特变换提取信号瞬时参数的方法。研究了经验模式分解与希耳伯特变换相结合的提取信号瞬时参数的EMD／HS法，并针对随机噪声的影响，提出了基于阈值的正交小波变换去噪法。理论分析及仿真结果表明，该法克服了直接运用EMD分解中由于大量噪声带来的不必要的干扰，减少了EMD的分解层数以及累积边界效应对信号分析的影响，提高了瞬时参数提取的时效性和准确性。

Instantaneous Parameters Extraction Based on Wavelet Denoising and EMD

In order to decrease the random noise influence upon the empirical mode decomposition(EMD),a method utilizing the empirical mode decomposition and Hilbert transfer after the wavelet denoising to extract instantaneous features of signals is described.For decreasing the influence of random noise,a threshold denoising method based on wavelet transform is proposed.Theoretic analysis and simulation results show that the method not only can extract instantaneous features,but also reduce the decomposition layers of signals and accumulative end errors on later decomposition.Compared with that using EMD directly,this method improves a lot in the speed of extraction and accuracy of the extracted parameters.