Research Of Brain-Computer Interface Used For Motor Control Based On Motor Imagery EEG

Brain-Computer Interface(BCI) system is a technology that used to realize thecommunication between human thoughts and outside objects bases on EEG signals,which can be instead of the brain’s normal output pathway of peripheral nerves andmuscles. BCI technology has been rapid development and applied in the military,medical, industrial, rehabilitation and control fields at home and abroad in theseyears. In the rehabilitation field, BCI technology can help to restore the abilities ofcommunication with others and environment control for those with severedisabilities.Based on some researches of national and international, This paper conductedseries of researches about BCI system based on motor imagery EEG signals. thecontents of the researches include experimental examples design,the EEG signalacquisition, features extraction, pattern classification, virtual reality etc, which weregot some effective results. In the experimental sample design aspect, This paper putsforward a method that imagine hand and leg which has color to play, through theexperiment proves that this method was improved compared to imagine nature handand leg movement in the various performance index.In view of the difficulty of extracting the obvious features of EEG signals and theflaw of the large dimensions of features in BCI, this paper used themulti-parameter common spatio-spectral pattern (MCSSP) to extract the features.The experiment verified that the MCSSP algorithm got more effective results inprocessing low dimensional datas compared with common spatial pattern (CSP) andcommon spatio-spectral pattern (CSSP) algorithms. On the other hand, this paperstudied four classification algorithm included ridge regression, regularization Fisherlinear discriminant analysis(FLDA), Bayesian linear discriminant analysis(BLDA),support vector machine(SVM) algorithm. these algorithms were applied in Graz datasets and the self-acquisition data sets for validation and comparison, improved thatthe BLDA classification algorithm had much more stability than three others.Secondly, this paper present a new detection method of idle state of asynchronous brain-computer interface system using the classification advantages of BLDAalgorithm.Based on the researches of key technologies which were mentioned above, thendesigned an on-line BCI system: realization of virtual reality vehicle directioncontrol through image playing of left and right hand which have color.The SYMTOPUEA-24BZ EEG amplifier used for EEG signals acquisition. using VC++MFCplatform to designed a software system to analysis and transmission datas, finallythrough the movement changes of the virtual reality scene car,the classificationresult was fed to the experimenter.