航天噪声环境对脑机接口的影响研究  

Influence of Cabin Noise on Brain-Computer Interface in Space

作　　者：张朕[1] 焦学军[1] 杨涵钧[1] 徐凤刚[1] 姜劲[1] 曹勇[1] 潘津津[1] 

机构地区：[1]中国航天员科研训练中心人因工程国家重点实验室,北京100094

出　　处：《载人航天》2017年第2期274-278,共5页Manned Spaceflight

基　　金：国家自然科学基金(81671861);中国航天医学工程预先研究项目(YJGF151204);中国航天员科研训练中心人因国家重点实验室自主课题(SYFD150051805)

摘　　要：针对目前研究较薄弱的脑机接口在舱内噪声环境下的适应性问题,通过调整天宫一号舱内噪声录音至60dB以完全模拟太空舱内环境噪声,探究了P300、SSVEP和运动想象三种脑机接口在模拟天宫一号舱内噪声环境下分类正确率和特征的变化。结果显示:模拟天宫一号噪声环境与安静环境(20~40dB)相比,P300靶刺激幅值发生了较显著提高,但是分类正确率无显著性改变;SSVEP的各频率信噪比和分类正确率无显著性改变;运动想象正确率下降且显著,但仍在适用范围。表明脑机接口在载人航天噪声环境下具有较强适用性。BCI （ Brain-Computer Interface） can control remote machines through brain signal which is of great significance for the human spaceflight. The study of BCI is usually conducted in silent en-vironment, but the noise in space environment （ 60 dB ） is almost in-avoidable. Since there is few study of BCI in noise environment, the feasibility of BCI in the noise environment of the space cabin was explored in this paper. The noise recording was used in the cabin of Tiangong-1 to simulated the noise environment in the space cabin and three kinds of BCIs：P300 , SSVEP and MI （ Motor Image-ry） were studied. The results showed that：for P300-BCI, the target amplitude increased significant-ly but the classification accuracy was not affected by noise;for SSVEP-BCI, the SNR and classifica-tion accuracy were not affected by noise;For MI, the classification accuracy decreased significantly but was still in the normal range. Thus, the feasibility of these three kinds of BCI in noise environ-ment of manned spaceflight was confirmed.

关 键 词：脑机接口 载人航天 噪声 P300 SSVEP 运动想象 

分 类 号：R318[医药卫生—生物医学工程]