基于混合阈值的清除重复间隔阈值经验模态分解去噪方法

经验模态分解(Empirical mode decomposition,EMD)被认为是一种有潜力的非线性非静态信号去噪方法。传统的经验模态分解阈值去噪在零点附近存在不连续性的缺点,Kopsinis提出了EMD-IIT和EMD-CIIT方法,但这两种方法对阈值过于敏感,即区间极值轻微的偏差就有可能导致去掉整个区间曲线,因此本文提出一种混合阈值算法,结合了EMD-DT和EMD-IT各自的优势。仿真结果表明此去噪算法具有较好的效果。

DENOISING METHODS BASED HYBRID THRESHOLDING EMPIRICAL MODE DECOMPOSITION CLEAR ITERATIVE INTERVAL THRESHOLDING

Empirical mode decomposition is considered as a potential nonlinear and non-static signal denoising method. The traditional empirical mode decomposition threshold denoising has the disadvantage of discontinuity near zero. Kopsinis proposed the iterative EMD interval-thresholding and clear iterative EMD interval-thresholding methods. However, the two methods are too sensitive to the threshold. That is, the slight deviation of the interval extreme value may lead to the removal of the entire interval curve. Therefore, a hybrid thresholding algorithm which combines the strengths of EMD-DT and EMD-IT methods is proposed in this paper. Simulation results show that the denoising algorithm has good results.