东北大学学报（自然科学版）

东北大学学报(自然科学版) ›› 2010, Vol. 31 ›› Issue (1): 12-15.DOI: -

• 论著 • 上一篇    下一篇

基于共空间模式和神经元网络的脑-机接口信号的识别

叶柠;孙宇舸;王旭;   

  1. 东北大学信息科学与工程学院;
  • 收稿日期:2013-06-20 修回日期:2013-06-20 出版日期:2010-01-15 发布日期:2013-06-20
  • 通讯作者: -
  • 作者简介:-
  • 基金资助:
    国家自然科学基金资助项目(50477015)

Recognition based on common spatial patterns and ANN for brain-computer interface signal

Ye, Ning (1); Sun, Yu-Ge (1); Wang, Xu (1)   

  1. (1) School of Information Science and Engineering, Northeastern University, Shenyang 110004, China
  • Received:2013-06-20 Revised:2013-06-20 Online:2010-01-15 Published:2013-06-20
  • Contact: Ye, N.
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

969

摘要： 提出了一种基于共空间模式和LVQ神经元网络对不同意识的脑电信号进行分类的方法.脑电信号是通过电极在头皮表面采集的脑-机接口的控制信号,提取脑电信号特征并对其进行分类,组成不依赖于正常的由外围神经和肌肉组成的输出通路的通讯系统.首先利用小波包分解对原始脑电信号进行预处理,对分解后特定小波包子带的脑电信号进行共空间模式分解,提取最优的特征;然后利用LVQ网络对不同意识任务特征进行分类,实验结果表明,该方法取得了92.7%的平均分类识别率,已经达到脑-机接口实际应用的标准.

关键词: 脑-机接口, 小波包子带, 脑电信号, 共空间模式, 学习矢量量化

Abstract: Classifies the EEG signals of different ideas based on the common spatial patterns(CSP) and LVQ-ANN. EEG is the control signals of brain-computer interface(BCI), which are collected from one's scalp by electrodes to extract and classify the EEG features so as to form a communication system that does not depend on the brain's normal output channels of peripheral nerves and muscles. With the original EEG signals preprocessed by wavelet packet decomposition, the CSP is introduced to decompose further the EEG signals from the specified subbands of wavelet packet so as to extract the best features and classify the features which are of different ideas by LVQ-ANN. Simulation result showed that the method proposed can provide a recognizable accuracy up to 92.7% in classification, it has come up to the standard for the practical application of BCI.

中图分类号: