一种改进的非线性PCA胎儿心电信号提取算法  

Fetal Electrocardiagram Extraction Based on Improved Nonlinear Principle Component Analysis Algorithm

作　　者：夏冰[1] 刘亚雷[2] 穆晓敏[1] 齐林[1] 杨守义[1] 

机构地区：[1]郑州大学信息工程学院,郑州450001 [2]西南交通大学电气工程学院,成都610031

出　　处：《郑州大学学报（理学版）》2009年第1期99-103,共5页Journal of Zhengzhou University:Natural Science Edition

摘　　要：传统非线性主分量分析(PCA)算法在提取胎儿心电信号(FECG)时,对信道参数变化的跟踪能力较差,为有效地跟踪信道参数的变化,快速提取胎儿心电信号,提出一种通过预测误差调整遗忘因子的非线性PCA算法.该算法具有收敛速度快,参数估计误差小的优点.仿真结果表明,该算法与传统算法相比,可以快速有效地抑制掉母亲心电信号(MECG)和各种噪声,提取更清晰的胎儿心电信号.The traditional nonlinear principle component analysis algorithm has bad tracking capability to the variable channel parameter.An improved nonlinear principle component analysis algorithm is proposed,which adjusts the variable forgetting factor by the estimated error.This algorithm has faster convergence and less parameter.The simulations show maternal electrocardiogram and noises can be more quickly and effectively restrained by the proposed algorithm in contrast with the traditional algorithm,and the effe...

关 键 词：非线性主分量分析 胎儿心电信号 母亲心电信号 盲信号分离 

分 类 号：R318[医药卫生—生物医学工程]