基于事件相关电位的BCI新型输入系统研究

Study on the new input of BCI system based on ERP

设计了基于事件相关电位(ERP)的脑-机接口(BCI)系统,根据ERP中P300的特性,分别对比输入字符的形式及颜色,并在此基础上采用不同的实验参数进行了多次实验。对采集到的脑电信号(EEG)使用叠加平均、独立成份分析(ICA)、Fisher线性分类器来进行分析处理。实验结果表明,输入方式及实验参数会对P300的时域特性及识别率产生较大的影响。

This paper designs a BCI （Brain-Computer Interface） system based on ERP （Event-Related Potentials）, this paper has carried on many times experiment by using different experimental parameters based on comparing of the input character＇s form and color, according to the P300 potential characteristics. The collected EEG（electroencephalogram） data can be analyzed and processed by average, ICA （Independent Component Analysis）, Fisher linear classifier. The experimental results show that it can produce great influence about the time domain and recognition rate of P300 characteristics by the input method and experimental parameters.