基于EEMD和WT的运动想象脑电信号消噪方法

采集到的运动想象脑电信号MI EEG(Motor Imagery Electroencephalogram)通常含有大量噪声信号.为了消除噪声同时保留尽可能多的有效信号,本文提出了将集合经验模态分解EEMD(Ensemble Empirical Mode Decomposition)与改进小波阈值法相结合的消噪方法.改进小波阈值法采用了新的阈值选取规则和阈值函数.首先对信号进行EEMD分解,然后再对高频固有模态函数IMF(Intrinsic Mode Functions)进行改进小波阈值处理,最后将处理后的高频IMF分量和低频IMF分量进行重构得到消噪信号.以信噪比和均方根误差作为消噪效果的定量评价指标,将本文提出的方法与单纯使用EEMD分解消噪法、单独使用改进小波阈值消噪法、EMD与改进小波阈值法相结合消噪法进行比较,结果表明,本文提出的消噪法优于其他三种消噪法.

De-noising method of the MI EEG based on EEMD and WT

In order to eliminate the noise mixed in Motor Imagery Electroencephalogram(MI EEG)and retain useful MI EEG information,the paper puts forward a new MI EEG de-noising method based on ensemble empirical mode decomposition(EEMD)and improved wavelet threshold method. New threshold function and threshold selection rules are introduced to the improved wavelet threshold denoising method. Firstly,the MI EEG signal is decomposed by the EEMD. Then using the improved wavelet threshold method to denoise the high-frequency Intrinsic Mode Function(IMF)components. Finally,the processed high frequency IMF components and low frequency IMF compo-nents are reconstructed to get the denoised signal. The experimental results reveal that the proposed de-noising algo-rithm has perspective of higher SNR and lower RMSE compared to the other methods,including the pure EEMD, the pure improved wavelet threshold method,and the improved wavelet threshold method based on EMD.