Research On Visual Brain-Computer Interface Based On Coding And Frequency Sequential Modulation

Brain computer interface(BCI)can help human to interact with external devices without relying on the peripheral nerves of brain and the muscle.It can extract the characteristics of human brain electrical signals to interpret the user’s intention and realize the function of controlling external devices.In recent years,BCI based on visual evoked potentials(VEP)has become a research hotspot in the field of brain computer interface.Compared with other BCI systems based on VEP,the SSVEP(steady state VEP)BCI and c-VEP(coded modulated VEP)BCI have more stimuli targets and higher information transmission rate.However,the hybrid modulation BCI system based on the two has not been further studied.In this paper,the hybrid BCI system based on SSVEP and c-VEP is studied,and a new hybrid BCI paradigm based on coding and frequency sequence modulation is proposed to realize a high-performance hybrid BCI system.In this paper,based on the research of traditional visual BCI with single EEG mode,a hybrid BCI system is designed by using pseudo-random code and frequency code to modulate visual stimulus target.Twelve volunteers were invited to participate in the experiment,and their hybridized EEG data were extracted and analyzed to verify the feasibility and effectiveness of the system.Firstly,the autocorrelation function and cross-correlation function of four c-VEP template signals and the spectrum of SSVEP template signals of one volunteer are analyzed.The results show that the c-VEP and SSVEP induced by the hybrid system have excellent separability.Then the classification recognition rate and information transmission rate of the system are analyzed through the test data.The results show that the average recognition rate of 64 targets in the system is 94.27%,and the information recognition rate is 105.07 bit/min,which proves that the hybrid BCI system has excellent recognition rate.Finally,the optimization analysis of the system is carried out from two aspects:the interaction between SSVEP and c-VEP and the influence of SSVEP data length on the system performance.This paper find that the data delay of 190ms can greatly reduce the interference between SSVEP and c-VEP and optimize the system performance;when the SSVEP data length is 0.5s,it can maintain a high recognition rate of 92.06%and increase the information transmission rate of the system to 120.87 bit/min.