【作者】 钟伯成；

【Author】 ZhongBoCheng Department of Computer Science & Technology,Hefei University,230022

【机构】 合肥学院计算机科学与技术系；

【摘要】 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,对基于信息极大原理的独立分量分析算法的动态盲分离进行了研究。仿真实验结果表明基于单次样本更新的在线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。更多还原

【Abstract】 For the limits on analyzing the strong non-stationary bio-electrical signals by the static independent component analysis (ICA)algorithm,the paper researched the dynamic blind source separation using the independent component analysis(ICA)algorithm based on informarion maximization.Simulation experiments show that the on line ICA algorithm has good performance in blind source separation,and also achieves excellent result with time variant mixing.更多还原