The Study For P300 Brian-computer Interface Based On OpenViBE Platform

Brain computer interface(BCI)is a special kind of human computer interaction system,which can detect the brain activity and convert it into command output.The realization of this interaction does not depend on the peripheral nerve and muscle tissue.In this paper,the brain computer interface based on P300 is studied.P300 is the most representative evoked potential in event-related potentials(ERP).The brain computer interface experiment,which is based on P300,implements without special training,and is conducive to the implementation.At the same time,it can be induced by auditory,visual,tactile and other ways to induce,it can broaden the scope of application of BCI technology.Based on the existing research of P300 experimental paradigm and classification recognition,this paper has carried out the following work.1.Through the combination of visual evoked potentials and auditory evoked potential,this paper proposes a hybrid model of stimulus presentation,P300 waveform corresponding this way better than a single way is more significant,and the use of classification and recognition method of linear discriminant classification(LDA),the correct classification rate has been significantly improved.2.Based on the traditional P300 RCF paradigm,this paper presents a mixed area separated single flash ASF stimulation paradigm,effectively overcomes the problem of neighborhood interference,adjacent interference and double flash error of traditional RCF,and the classification rate and the rate of information transmission has obvious progress.3.In order to further improve the information transmission rate under the ASF paradigm,this paper proposes an adaptive algorithm based on the characteristics of the ASF paradigm.The adaptive algorithm controls the number of superimposed averages in the feature extraction process by setting the threshold in advance.The information transfer rate of the adaptive algorithm has increased by about 4bit/min on the original basis.4.In this paper,OpenViBE+MATLAB is used as the software platform of brain computer interface experiment.However,due to the aggregation of its function,OpenViBE is lack of flexibility in the experimental design.Its function can not meet the stimulation of the improved ASF paradigm.In this paper,the communication mechanism between OpenViBE and MATLAB is studied deeply,using TCP protocol to solve the signal synchronization problem of stimulus presentation..It effectively solves the flexibility of P300 experiment design and synchronization of EEG signal and stimulus signal.The improved experiment is realized.