Study And Implementation Of BCI System Based On Embedded System

Brain-computer interface(brain-computer-interface,BCI)connects the brain or nervous system to any processing or computing device through the intermediary of information exchange,so that the brain can control machine and machine can read brain's information.At present,the research of BCI system is still in the laboratory stage,and the acquisition equipment of EEG signal is usually complex and huge,therefore,this paper aims to research and implement a small-size,low-power BCI system based on embedded platform.Firstly,this paper summarized the characteristics of brain-computer interface technology and EEG signal,then analysed the processing method of the EEG signal and the principle of SSVEP-BCI system.The EEG signal is weak,high noise,non-stationary,nonlinear and random signal,and it is low frequency signal.Then,the stimulator and acquisition device in SSVEP-BCI system was studied.Because LED is not affected by refresh frequency,it can produce a large number of effective stimulus.Therefore,this paper designed a stimulator based on LED.Based on ADS1299 bioelectric potential acquisition chip,AVR MCU and Bluetooth chip,a kind of EEG collector is constructed with a small size and low power.In this device,ads1299 is the front-end acquisition module to complete the analog-to-digital conversion of EEG signal.As the main control module,AVR reads the data of ADS1299 through SPI and sends it to Bluetooth chip,which can transmit the collected EEG data to the processing device.Next,this research used matlab to study the EEG signal processing method,and used the Qt platform to complete the EEG data processing on the embedded development board.The original SSVEP signal was filtered by the 5Hz~30Hz Butterworth bandpass filter,and the processing of the ssvep signal was carried out by using the FFT algorithm and the CCA algorithm.The EEG signal processing was implemented based on Qt platform.The software can complete the tasks of receiving data with serial port,displaying real-time waveform and extracting the frequency of ssvep signal.Finally,the accuracy of the stimulator was measured by oscilloscope,and the braincomputer interface experiment of ssvep was designed to verify the overall function of the system.The 5 stimulus frequency was chosen as 7Hz,7.5Hz,8.7Hz,11 Hz,and 12.7hz to verify the accuracy of the algorithm.The results of the experimental shown that the EEG signal collector designed in this paper can collect high quality EEG signals,and the processing software can accurately obtain the processing results,thus the effectiveness of the whole system was validated.