摘要
文章针对静态独立分量分析(ICA)算法在分析非平稳性很强的生物电信号方面的局限,研究了基于信息极大原理的动态独立分量分析算法,并将它与静态ICA算法进行了比较;仿真实验结果表明,动态ICA算法具有较好的盲源分离性能,并且能够在时变混合情况下,获得较好的盲分离效果。
For the limitations of the static independent component analysis(I CA) algorithm in analyzing the strongly non-stationary bioelectrical signals, the paper studies the dynamic independent component analysis algorithm based on information maximization and compares it with the static ICA algorithm. Simulation results show that the dynamic ICA algorithm has good performance in blind source separation, and also achieves an excellent result with time variant mixing.
出处
《合肥工业大学学报(自然科学版)》
CAS
CSCD
北大核心
2009年第8期1154-1157,共4页
Journal of Hefei University of Technology:Natural Science
基金
合肥学院人才基金资助项目(600840)
关键词
动态ICA算法
信息极大
盲源分离
dynamic ICA algorithm
information maximization
blind source separation(BSS)