摘要
基于LabWindows/CVI和Matlab设计一个BCI在线控制系统,对8Hz～30Hz的运动想象脑电信号提取时域均值、中值偏差估计、瞬时能量均值、AR模型参数等特征,应用增量式支持向量机进行分类,实现人脑对虚拟汽车直接控制。系统采用了多线程技术,保证各项工作的同时进行,在CVI中完成脑电数据采集、Matlab调用和控制指令的发送,在Matlab中进行脑电模式识别,两个程序共同完成对虚拟汽车的运动控制。经过实际测试证明,该系统具有操作简单方便、界面友好、可扩展性强、效率和可靠性高等优点,进一步推动了BCI的应用。
Based on LabWindows/CVI and Matlab,a BCI online control system is designed to extract time-domain mean,median deviation estimation,instantaneous energy mean,AR model parameters and other features from 8Hz^30Hz EEG signals and apply incremental support vector machine to classify so as to enable the human brain to directly control a virtual vehicle.The system adopts multi-thread technology to ensure different works are simultaneously processed to complete EEG data acquisition,Matlab calls and control instructions sending in CVI,and carry out EEG pattern recognition in Matlab to enable two programs collaboratively control the motion of a virtual vehicle.After actual experiment,it is proven that the system bears such advantages as simple and convenient operation,friendly interface,high scalability,high efficiency and reliability etc.,thus further promotes the application of BCI.
出处
《计算机应用与软件》
CSCD
北大核心
2012年第5期16-19,共4页
Computer Applications and Software
基金
国家自然科学基金项目(60975079)
上海市教育委员会创新项目(11YZ19)
