独立分量分析和遗传算法相结合的运动想象频带优化  

Band optimization of motor imagery based on genetic algorithm and independent component analysis

作　　者：康莎莎 周蚌艳[1,2] 吴小培[1,2] KANG Shasha;ZHOU Bangyan;WU Xiaopei(School of Computer Science and Technology,Anhui University,Hefei 230601,China;The Key Laboratory of Intelligent Computing and Signal Processing,Anhui University,Hefei 230039,China)

机构地区：[1]安徽大学计算机科学与技术学院,安徽合肥230601 [2]安徽大学计算智能与信号处理教育部重点实验室,安徽合肥230039

出　　处：《安徽大学学报（自然科学版）》2019年第1期37-44,共8页Journal of Anhui University(Natural Science Edition)

基　　金：国家自然科学基金资助项目(61271352,61401002)

摘　　要：对结合独立分量分析(independent component analysis,简称ICA)和遗传算法(genetic algorithm,简称GA)的运动想象脑电(motor imagery electroencephalogram,简称MI-EEG)特征检测及其优化方法开展研究.设计了基于ICA的MI-EEG分类算法.在此基础上,针对不同受试个体,用GA算法对运动想象诱发的事件相关去同步(event-related desynchronization,简称ERD)频段进行优化选择,用以改善运动想象脑一机接口(brain-computer interface,简称BCI)系统的识别率.实验结果表明,基于ICA的GA算法特征优化方法具有较好的可靠性和实用性,可用于在线BCI的设计与实现.The paper focused on the research of feature detection and the optimization algorithm of motor imagery electroencephalogram (MI-EEG)by using independent component analysis (ICA)and genetic algorithm(GA),and a MI-EEG classification method based on ICA was proposed in this paper.For different individual,the method used GA to optimize selection of the optimal event-related desynchronization(ERD)band which was induced by motor imagery,and it improved the recognition rate of motor imagery brain-computer interface (BCI)system.Experimental results revealed that the feature optimization of GA method based on ICA,which not only had better reliability and practicability,but could be used for the design and implementation of online BCI.

关 键 词：脑-机接口 脑电信号 运动想象 遗传算法 独立分量分析 节律增强频带 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]