A MI-SSVEP Hybrid Brain Computer Interface System for Online Control of a Mobile Vehicle

Brain-Computer Interface (BCI) is a device that translates intentionally modulated brain signals into control commands. Based on motor imagery (MI) and steady-state visual evoked potentials (SSVEP) recorded from human subjects, motor commands could be decoded and translated to achieve the desired movements. However, achieving high levels of accuracy and stability in online BCI remains a challenge that needs to be addressed. This study utilized a limited number of channels (C3, C4, O1, and O2) to extract EEG frequency domain features through continuous wavelet transform. Furthermore, the classifiers of Bayes, KNN, and SVM were employed to decode motion commands. Moreover, a majority voting strategy was adopted to improve the online accuracy of the proposed classification model. The research results indicated that incorporating the majority voting algorithm into the online BCI control process could improve the average accuracy of mobile vehicle direction control from 68.33% to 87.08%.