| Electroencephalography is an electrophysiological signal result from human neurons,reflecting individuals’ cognitive activities.The brain-computer interface is a communication system between human consciousness and the outside world.Steady-State Visual Evoked Potential can be acquired by gazing at the flicker with a certain frequency on the screen.Because of the simple structure,less training proceeding and high information transfer rate,SSVEP-based BCI system has been widely applied in various EEG researches.This thesis firstly design SSVEP-based stimulus on the LCD screen in Java language and succeed in inducing subject’s EEG signal.Focusing on the inconvenience of the BCI system,then a cheap EEG acquicition system in small size has been built with ADS1299 biochip.The original signal is filtered through a conditioning cuircuit and amplified to a monitorible degree.And then transfer to digital signal through ADS1299.The modulated-signal is finally displayed on the screen by RFD22301.As a result,the designed-system has abtained subject’s EEG signal sucessfully with high resolution ratio.Additionally,Independent Component Correlation Algorithm is applied to reduce redundant artifact in muti-channels.Fast-fft and Canonical Correlation Analysis is used in classification proceeding.Comparing the classification results in two methods,it is suggested that CCA has a better performance than FFT whenever in a short time window or a longer one.At last,according the portability of openBCI,a convenient anti-interference BCI system has been invented to control a smart car.In online experiment,all the four subjects have achieved such a higher average classfication rate 92.5% with a further improvement among most BCI-online controlling systems.This experiment also comfirmed the instantaneity and efficiency of this intelligent car controlling system based on ICA and CCA,providing a novel solution to advance the robustness and accuracy in online controlling systems. |
PDF Full Text Request