摘要
为消除胃动力阻抗信号中混叠的噪声信号,利用独立分量分析的冗余取消特性,提出一种新的胃动力阻抗信号消噪方法。采用经验模态分解构造虚拟噪声通道,将一维原始胃动力阻抗信号扩展为多维观测信号,应用FastICA算法对其实施盲分离。仿真实验结果表明,该方法能有效消除叠加在胃动力阻抗信号中的噪声,不需要大量的观测样本,可运用独立分量分析实现对单个观测样本的消噪处理。
In order to eliminate aliasing noise signal of gastric motility impedance signals,using independent component analysis of the redundancy reduction features,a new method of gastric motility impedance signals denoise is proposed.Construction of virtual noise channel used empirical mode decomposition,the one-dimensional original gastric motility impedance signal is extended to multi-dimensional observation signals,and it is implemented blind separate by FastlCA algorithm.Simulation experimental results show that the denoise method can effectively eliminate the noise of gastric motility impedance signal.The method does not require a large number of observation samples,which can achieve denoise of a single observed sample by Independent Component Analysis（ICA）.
出处
《计算机工程》
CAS
CSCD
北大核心
2011年第10期260-262,共3页
Computer Engineering
基金
重庆市渝中区科委课题基金资助项目"基于数据挖掘技术的HIS系统的研究"
