基于EMD自相关的表面肌电信号消噪方法

发布时间：2018-03-12 11:26

本文选题：表面肌电信号　切入点：经验模态分解　出处：《仪器仪表学报》2014年11期 　论文类型：期刊论文

【摘要】：为了更好地消除混杂在表面肌电信号(s EMG)中的噪声,提出一种基于噪声统计特性的EMD自相关消噪方法。首先,对含噪s EMG信号进行EMD分解,并根据噪声统计特性降低低信噪比的高频IMF分量的能量后重组信号。其次,对重组后的信号进行自相关函数特性的EMD分解,并对自相关函数方差低于阈值的高频IMF分量进行小波去噪。最后,把处理后的高频IMF分量和低频IMF分量重构,得到的信号即为消噪信号。实验结果表明,该方法不仅能更好的消除噪声,而且在低信噪比情况下有良好表现。
[Abstract]:In order to better eliminate the noise mixed in the surface EMG signal, a EMD autocorrelation denoising method based on the statistical characteristics of noise is proposed. Firstly, the noisy s EMG signal is decomposed by EMD. According to the statistical characteristics of noise, the energy recombination signal of high frequency IMF component with low SNR is reduced. Secondly, the EMD decomposition of autocorrelation function is applied to the reconstructed signal. The high frequency IMF component whose variance of autocorrelation function is lower than the threshold value is de-noised by wavelet transform. Finally, the processed high frequency IMF component and the low frequency IMF component are reconstructed and the signal is de-noised. The experimental results show that, This method not only eliminates noise better, but also performs well under low SNR.
【作者单位】：杭州电子科技大学智能控制与机器人研究所;
【基金】：国家自然科学基金(60903084,61172134) 浙江省自然科学基金(LY13F030017) 浙江省科技计划(2014C33105)项目资助
【分类号】：TN911.4

【相似文献】