一种新的EMD去噪方法

A New Signal Denoising Method Based on Empirical Mode Decomposition(EMD)

信号的经验模态分解方法(EMD)可以理解为以信号极值特征尺度为度量的时空滤波过程。此过程充分保留了信号本身的非线性和非平稳特征,在信号的滤波和去噪中具有很大的优势。文中分析了EMD去噪技术的基本原理,通过研究几种利用EMD去除非平稳信号噪声的技术,提出了一种新方法。仿真结果表明新的EMD去噪法在去除噪声,提高信噪比方面具有极大的优越性。

The empirical mode decomposition （EMD） method of signal can be interpreted as time and spatial filtering based on the signal＇s extremum eharaeteristic seale. The proeess preserves the nonlinear and nonstationary eharaeteristies of the signal whieh is a great advantage in filtering and denoising. The paper analyzes the basie theory of removing noises from signals with EMD, and researehes several methods using the technique of EMD to remove noises from the unsteady signals, and then proposes a new method. Simulation results indieate that the new EMD denoising method has great advantages in denoising and raising SNR.