CAJ | 학술논문

针对传统小波阈值去噪算法中阈值选取困难，提出一种新的小波-EMD-SVD差分谱组合模式。对原始信号做小波分解得到一系列细节信号，不再通过小波阈值将强噪声小波系数简单地置零，而是对其进行EMD-SVD差分谱微弱信号提取，最后进行重构。实验结果表明，该算法对强噪声环境下信号的去噪效果明显且平稳。
침대전통소파역치거조산법중역치선취곤난，제출일충신적소파-EMD-SVD차분보조합모식。대원시신호주소파분해득도일계렬세절신호，불재통과소파역치장강조성소파계수간단지치령，이시대기진행EMD-SVD차분보미약신호제취，최후진행중구。실험결과표명，해산법대강조성배경하신호적거조효과명현차평은。
For the difficulty of selecting threshold value in the traditional wavelet threshold denoising method,a new type of wavelet-empirical model decomposition ( EMD) -singular value decomposition ( SVD) is proposed. An original signal is decomposed by wavelet transform and a group of detail signals are obtained, however, heavy noise wavelet coefficients are not simply set to zero by wavelet threshold any more, whereas these heavy noise wavelet coefficients are used to EMD and SVD difference spectrum decomposing and extracts the weak signal. Finally reconstructs the single. The result indicates that the denoising effects are obvious and smooth in heavy noise surrounding.