Study On The Performance Of Brain-Computer Interface Based On Steady-State Visual Evoked Potential

In recent years,with the continuous progress of science and technology,the development of brain computer interface(BCI)system has also ushered in an accelerated stage.Among them,the brain computer interface based on steady state visual evoked potentials(SSVEP)has been widely concerned.The brain computer interface system based on SSVEP signal has the characteristics of simple use,less training time,high information transfer rate(ITR)and stable effect.It can generate characteristic and rhythmic signals in the cerebral cortex through external visual stimulation,with good stability and continuity,and can be used for signal recognition and calculation.Because the power spectrum component of SSVEP signal at the excitation frequency is very strong,important features can be extracted by power spectrum analysis and canonical correlation analysis(CCA),which makes SSVEP signal widely used in BCI system.However,for BCI system,the accuracy,real-time and stability of the signal need to be guaranteed in order to ensure the normal operation of the system,so this puts forward higher requirements for the quality of SSVEP signal.How to accurately analyze the characteristics of SSVEP signal online and design a real-time and stable online BCI system based on SSVEP signal is one of the key contents of current BCI system research.How to choose the method of target recognition and what parameters to set have a very important influence on the result of target recognition.Therefore,it is of great significance for the research and development of BCI system to compare the advantages and disadvantages of different algorithms.This paper summarizes the acquisition methods of SSVEP signal,studies the factors that affect the signal quality,and summarizes the methods to obtain better signal from the aspects of external equipment,light frequency and light color that induce visual response.Secondly,through off-line analysis of EEG signals with different methods,by comparing the recognition effect of different methods,this paper puts forward the advantages of likelihood ratio test method,and determines the feasibility of likelihood ratio test in practical application.In addition,this paper focuses on the performance comparison between different algorithms based on CCA,mainly from the accuracy of the algorithm,information transmission rate and the running time of the algorithm.By comparing the results,the characteristics of different algorithms are summarized,which provides a reference for the selection of algorithm in practical application,and provides some guidance for the improvement direction of the algorithm in the future.