摘要
[目的/意义]脑电信号出现的特征分类精确度较低、视觉刺激器稳定性和安全性不高的问题亟待解决,以提高脑机接口系统的广泛应用前景.[方法/过程]设计了一种基于稳态视觉诱发电位(SSVEP)脑控机器人的指令识别方法,运用Pygame图形技术设计的箭头棋盘视觉刺激器,刺激受试者大脑产生稳SSVEP信号数据,然后通过滤波器组典型相关分析(FBCCA)的方法,对SSVEP信号识别和分类,将分类后的控制命令结果数据发送给机器人,控制完成相应动作.[结果/结论]通过对6名受试者的实验结果显示,与传统视觉刺激器相比,箭头棋盘视觉刺激器提高了视觉刺激器稳定性、安全性和舒适度.这套系统的最高控制准确率和平均信息传输率,分别达到90.91%和20.98 bit/min,提高了特征分类数据的精确度、控制准确率.
[Purpose/Significance]The problems of low accuracy of EEG feature classification and low stability and safety of visual stimulator need to be solved urgently,in order to improve the practical application prospect of brain computer interface system.[Method/Process]This paper designs an instruction recognition method based on SSVEP brain-controlled robot.The Pygame graphics technology is employed to design an arrow checkerboard visual stimulator.The subjects'brains were stimulated to generate steady-state Visual Evoked Potential(SSVEP)signal data,and then the SSVEP signal was classified by filter bank canonical correlation analysis(FBCCA).The classified control command result data is sent to the robot to control its movement to complete the corresponding action.[Results/Conclusion]The experiments were conducted on six subjects.Compared with traditional visual stimulator,the experimental results show that the arrow checkerboard visual stimulator improved the stability,safety and comfort of the visual stimulator,and the maximum control accuracy and average information transmission rate of the system reached 90.91%and 20.98 bit/min,respectively,which improve the precision of the feature classification data,control accuracy.
作者
王逢生
谭平
苏晓萌
邱辉
Wang Fengsheng;Tan Ping;Su Xiaomeng;Qiu Hui(School of Intelligent Engineering and Intelligent Manufacturing,Hunan University of Technology and Business,Hunan Changsha 410205;Hunan Xiaodian Automobile Service Co.,Ltd.,Hunan Changsha 410006)
出处
《网络空间安全》
2022年第5期41-45,共5页
Cyberspace Security
基金
湖南省研究生科研创新项目资助(项目编号:CX20211152)。
关键词
脑机接口
机器人
稳态视觉诱发电位
滤波器组典型相关分析
数据安全治理
brain-computer interface
robot
steady-state visual evoked potential
filter banks canonical correlation analysis
data security governance
作者简介
王逢生(1997-),男,汉族,山东济宁人,湖南工商大学智能工程与智能制造学院,在读硕士,主要研究方向和关注领域:脑机接口、人机交互;谭平(1981-),男,汉族,湖南涟源人,中南大学,博士,湖南工商大学智能工程与智能制造学院,副教授,主要研究方向和关注领域:脑机接口、模式识别和信号处理;苏晓萌(1997-),女,汉族,山东德州人,湖南工商大学智能工程与智能制造学院,在读硕士,主要研究方向和关注领域:脑机接口、人机交互;邱辉(1980-),汉族,湖南长沙人,湖南大学,硕士,主要研究方向和关注领域:智能信息处理。
