Brain Computer Interface Based On P300 And SSVEP Hybrid Paradigm

EEG(Electroencephalogram)is the acquisition of physiological electrical activity of brain cells on the scalp surface,which is closely related to the activity state of human consciousness.Through the analysis of the EEG signal,can identify different awareness activities,use identify consciousness as a control signal to form a control system that does not depend on the brain nervous system and muscular system can communicate,known as brain computer interface(Brain-Computer,Interface,BCI)system.P300 is a kind of potential and the attention of endogenous event related evoked potentials,is a kind of psychological or language in special potentials,which mainly relies on the reaction and induced events do not depend on the person's physical condition.On the graphics or flashing lights and other visual stimuli,the brain cortex of the occipital nerve cell spontaneous electrical activity of the brain specific response to stimulation is called visual evoked potential(Visual Evoked,Potential,VEP).According to the frequency of flicker stimulation,the evoked potential is different.Transient VEP(TSVEP)and Steady State VEP(SSVEP)are generally divided into two categories.The mixed paradigm system can be either a serial completion of the two paradigms or a parallel connection.The serial mixed paradigm of brain computer interface systems,the general preceding paradigm,acts as a "brain switch"(brain switch).That is,the system can process two kinds of signals simultaneously.Obviously,the parallel brain computer interface paradigm requires simultaneous recording of data signals,information fusion,and model design,which greatly increases the complexity of brain computer interfaces.This paper focuses on the paradigm of brain computer interface stimulation in P300.Firstly,the typical oddball paradigm of P300 and the SD stimulus paradigm are introduced,and three linear classification methods are analyzed.Mainly introduces two kinds of stimulation paradigm(oddball paradigm and SD paradigm)respectively in three FLDA,BLDA,SWLDA classification algorithm,the classification results obtained in the three classification mode,each participant P300 waveform characteristics produced by different personnel,but roughly the same.Compared with the two paradigms,the SD stimulus paradigm is more likely to stimulate P300 signals.Compared with the three pattern recognition algorithms,SWLDA has the best classification accuracy and time effect,and is the classification algorithm used in the mixed paradigm in this paper.The SD stimulus paradigm is also the choice paradigm of the P300 paradigm in this paper.In this paper,we mainly explore the SSVEP paradigm,and introduce the current mainstream SSVEP stimulus paradigm,namely Newton ring stimulation and sinusoidal encoding stimulus.SSVEP color stimulus paradigm and SSVEP duty cycle stimulus paradigm are mainly studied.FFT,CCA and FBCCA three are used to study the algorithm respectively.The feature value is extracted to select the best color and duty cycle stimulus paradigm.The analysis results show that the system performs best when the green duty ratio is 50%.It is believed that the research results in this chapter will be helpful to the construction of high performance and accurate stimulus paradigm.By combining the optimal stimulus paradigm and the optimal processing method of P300 and SSVEP,the optimal mixed paradigm is constructed.A hybrid real-time brain computer interface system was developed on the Matlab platform to achieve the input system with 26 characters and 10 digits.The highest accuracy rate was 90%,and the average accuracy rate was 83%.