期刊
出版社
WORLD SCIENTIFIC PUBL CO PTE LTD
DOI: 10.1142/S0218001419590304
关键词
Motor imagery; multivariate empirical mode decomposition; common space pattern
资金
- key technologies of intelligent manufacturing integration for large scale assembly line of construction machinery of the key research and development plan of Shandong province [2016ZDJS02A12]
- research and application of key technologies for intelligent assembly data acquisition and processing of construction machinery based on industrial Internet of things of the major scientific and technological innovation projects in Shandong [2017CXGC0603]
- other research and application of key technologies [2018GGX103042, 2018YFJH0306, 2017CXGC0918, 2017CXGC1505]
The high accuracy of electroencephalogram (EEG) signal classification is the premise for the wide application of brain computer interface (BCI). In this paper, a hybrid method consisting of multivariate empirical mode decomposition (MEMD) and common space pattern (CSP) is proposed to recognize left-hand and right-hand hypothetical motion from EEG signals. Experiments were carried out using the BCI competition II imagery database. EEG signals were decomposed into multiple intrinsic mode functions (IMFs) by MEMD. The IMF functions with high correlation were processed by CSP, and AR coefficients and entropy values were extracted as features. After genetic algorithm optimization, classification is carried out. Our research results show that the K nearest neighbor (KNN) as an optimal classification model produces 85.36% accuracy. We also compare the proposed algorithm with the existing algorithms. The experimental results show that the performance of the proposed algorithm is comparable to or better than that of many existing algorithms.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据