Research On Portable SSVEP-BCI System

Brain-computer interface(BCI) is a communication control system using EEG signals to communicate with the outside world or to control external devices without any physical mechanical motion. A BCI system establishes a new channel for communication between brain and external environment. BCI systems based on the steady-state visual evoked potential(SSVEP) use visual stimulations to issue orders. These systems are widely used in laboratory studies and clinical trials, for the simplification of signal acquisition and nearly no training. This thesis aims to explore the portable SSVEP-BCI system with the merits of high mobility, low power consumption, to acquire high signal-to-noise ratio(SNR) EEG signals as well as extract SSVEP features correctly.This thesis studies and designs three important parts of the proposed SSVEP-BCI system:(1) Visual stimulator. By studying the requirements of SSVEP-BCI system for visual stimuli and analyzing the influences of various parameters on the visual stimulation effect, select the stimuli generation method reasonably, configure the parameters appropriately, and design the hardware circuit of the stimulator finally. Comparing the advantages and disadvantages of various modulation methods, the frequency modulation method is chosen and used ultimately to produce stable, accurate visual stimulations. During the stimulation frequencies’ accuracy test, the maximum frequency deviation is as low as 0.01 Hz.(2) EEG signal recording device. The design requirements are put forward through the analysis of the characteristics of EEG signals. Select the type and locations of the electrodes by studying the effects of electrodes on the SSVEP response. According to the design requirements, analyze the performance of similar devices or methods, then chose the core components and data transmission method, and finally design the EEG signal acquisition hardware circuit which mainly contains the analog-to-digital conversion(ADC), communication control, data transmission and power management module. The entire device is small in size(7×5cm), low power consumption(powered by 3000 mAH lithium battery can continuously record EEG for more than 30 hours).(3) Digital signal processing platform. Analysis the merits and defects about the operation systems deployed on the smart devices contributes to the decision of using Android based smartphone for digital signal processing. The application program is developed for digital EEG signal receiving, data storage, digital signal processing and SSVEP feature extraction.Select 5 volunteers to test the performance of the system. The average characteristic identification accuracy is more than 85% with the length of the recorded EEG is eight seconds. And the accuracy is higher than 90% when signal length is 40 seconds.Due to the advantages of high accuracy, small size, low power consumption and high mobility, the proposed portable SSVEP-BCI system is promising to be applied in many fields such as telephone communication and intelligent wheelchair control.