The Research For Key Technology Of Brain-computer Interface Based On Motor Imagery

BCI system is a communication system for controlling a device, e.g. computer, or wheelchair, by human intensions, which does not depend on the brain's normal output pathways of peripheral nerves and muscles but relies on the detectable signals representing responsive or intentional brain activities. BCI system use computer and other equipment to gather and analyze EEG data under specific environment and tasks, then the brain information is translated into control commands to achieve the purpose of controlling external devices. In this way, BCI provides a new communication channel that system developers can use in a variety of applications, such as assisting people with severe motor disabilities; supporting biofeedback training in people suffering from epilepsy, stroke; or controlling computer games. Nowadays, BCI technology has aroused the attention of many scientists from wide fields and has become a new field in the cross-over field which contains neural science, biomedical engineering, computer science and electronic information, and has high value in the fields of science and practice.Based on the knowledge of national and international BCI development, this paper studied the BCI system that was based on the left-right hand motor imagery from the online application and offline analysis. Concrete work is shown as follows:(1) The EEG acquisition system is established based on VC 6.0 software. In this system, The EEG data of visual evoked potentials and somatosensory evoked potentials can be collected, stored, and on-line analysis of power spectrum and EEG map is available.(2) EEG feature extraction. It is very critical to obtain a meaningful EEG feature which contains the related information of different mental actions. Feature extraction. The thesis designs experiment program based on motor imagery to acquire EEG. In the experiment, the subject is asked to imagine left or right hand movements according to the direction of the arrow. Independent Component Analyze (ICA) is used to remove the artifacts. With this method, the artificial caused by eye movements can be restrained effectively. The power spectrum is used in our research to obtain the EEG feature.(3) The classification system for color balls was designed based on left-right hand motor imagery. The system recognizes obtained EEG signal with the Fisher LDA method, then control the falling color ball into the appropriate container. The BCI system based on left-right hand motor imagery was elementarily realized.(4) Considering that the EEG signal is nonstationarity and nonlinear, a new-method based on wavelet packet entropy and support vector machines is applied for EEG signal recognition in this paper. The time-frequency characteristics and the uncertainty of EEG signal are analyzed by wavelet packet entropy, and extracts features of motor imagery from single trail. Then the feature data are classified by the support vector machines (SVM), meanwhile a finding-minimum-search method is proposed to determine the kernel parameter v and trade-off parameter c. Finally, some evaluation criteria including mutual information (MI), Signal-to-Noise Ratio (SNR) and misclassification rate (MR) are utilized to analyze the performance of classifier. The test results achieve that the recognition accuracy is 90%, Analysis has shown that WEP-SVM was an effective method for classification of the mental tasks. It can be applied in BCI system.This paper integrates some technologies including EEG analysis, motor imagery, signal processing algorithms and Brain-Computer Interface and so on, realizes the BCI system based on left-right hand motor imagery elementarily, lays solid foundation for application-oriented BCI system.