摘要
脑机交互的关键技术是脑电波(EEG)信号的解码,而人的大脑信号具有随时间改变的非稳态本质,探索EEG相对稳定的优秀特征表达非常关键。基于此,通过小波包分解、自适应带通滤波共空间模式、时-频-空多模态分析等人工智能等算法,提高特征表达鲁棒性。
The key technology of brain-computer interaction is EEG signals decoding,and human brain signals have an unsteady nature that changes with time,so it is very important to explore the relatively stable excellent feature representation of EEG.By means of wavelet packet decomposition,adaptive bandpass filter common spatial pattern,time-frequency-space multimode analysis and other artificial intelligence algorithms,the robustness of feature representation is improved.
作者
战中才
ZHAN Zhongcai(Shandong Taishan Academy of Sciences,Taian 271000)
出处
《现代制造技术与装备》
2020年第11期82-83,共2页
Modern Manufacturing Technology and Equipment
关键词
脑机交互
EEG
人工智能
鲁棒性
brain-computer interaction
EEG
artificial intelligence
robustness