摘要
提出了应用改进的非线性自适应多项式滤波器用于去除ECG信号噪声的方法。利用误差准则确定的Volterra序列展开能很好模拟生物信号成份,而基于Volterra序列的自适应滤波器能有效模拟ECG成份从而达到去除信号中噪声的目的。该方法在研究Volterra序列的特性并确定适应心电信号的多项式阶次、多项式结构等参数基础上,提出了改进的非线性多项式自适应滤波算法。针对MIT-BIH数据库数据的计算机模拟结果显示该方法能有效提高ECG的信噪比。
A modified nonlinear adaptive polynomial filter is proposed to reduce random noise in EGG signals. A truncated Volterra series expansion, whose polynomial number can be determined by error criteria, was used to model the biomedical signal component. Based on Volterra we proposed an adaptive filter that could approximate and extract the signal in noise corrupted EGG data efficiently. Firstly we investigated the characters of Volterra series so as to determine parameters of the polynomial, such as the dynamic order and the structure of Volterra series that fit the EGG signal. Secondly we presented the algorithm of nonlinear modified polynomial adaptive filter. Successful results with computer simulations on MIT-BIH database illustrated this method appropriate to enhance SNR of EGG.
出处
《北京生物医学工程》
2007年第1期57-59,共3页
Beijing Biomedical Engineering
