Font Size: a A A

Research On Brain - Computer Interaction Pattern Recognition Method Based On Motion Imagery EEG

Posted on:2016-12-25Degree:MasterType:Thesis
Country:ChinaCandidate:H W SunFull Text:PDF
GTID:2208330470468084Subject:Detection Technology and Automation
Abstract/Summary:PDF Full Text Request
The brain machine interface is a revolutionary human computer interaction. The brain computer interface based on motor imagery EEG is a very important kind of brain computer interface. Brain computer interface technology is a cross research involves many disciplines. Brain computer interface can be directly reconstruction of motion control by EEG. It can be strategically used for military purposes, at the same time can provide auxiliary control for the disabled and normal people, so as to improve the quality of life. So far, motor imagery EEG brain machine interface is still not out of the laboratory, practical application.This paper mainly used motor imagrey EEG as the research object. First designed specifically signal acquisition helmet from motor sensory area of cerebral. And studied the motor imagery EEG signal preprocessing, feature extraction and pattern recognition algorithm. Finally preliminarily established online brain computer interface system. The main contents of this thesis are as follows:mainly in the following several aspects to study, and achieved certain results:(1) Motor imagery acquisition helmet is Designed specifically collect EEG signal for the motor sensory cortex. Using sliding design, can better adjust the position of the electrode on the collected signal low impedance, better suitable for different sizes of the person wearing the head. Using dry electrodes which is more convenient used in our daily life. This helmet is expected in real life collection of motor imagery EEG signal, and can be used to study brain function cognition and brain controlled robot interface technology.(2) This paper studied the method for extracting motor imagery EEG features based on the Hilbert - Huang Bianhuan (HHT).According to the motor imagery EEG feature is nonstationary and nonlinear. HHT is used,which has a high resolution both in time domain and frequency domain. Then extracted AutoRegressive (AR) parameters and calculate the average instantaneous energy of motor imagery. Thus structural feature vector. Finally used support vector machine (SVM) to classifiy.SVM can well adapt to motor imagery EEG classification. This study confirmed that the HHT has good feature extraction ability to motor imagery EEG which is nonstationary and nonlinear signal. Also again confirmed that the ERD phenomenon of motor imagery. This paper can lay a solid foundation for research of online real-time brain-computer interaction control system based on motor imagery.(3) To explore the classification method of pattern recognition can be effectively applied to the left and right hand motor imagery EEG. Pattern recognition is an important part of brain computer interface system, a good classifier can improve the performance of brain computer interface system. This paper introduced BP neural network, naive Bias, linear discriminant analysis, support vector machine classifier, and the experiments were obtained from the rate of correct classification. To verify the effectiveness of these classifiers thought pattern recognition in motor imagery, and provides classification established for brain computer interface of the real-time brain machine control system.(4) Based on the above research experiences and achievements, This paper initial established the online motor imagery EEG BCI interactive control system.Which can lays the foundation for development of practical online real-time brain computer interface system,and accumulate experience.
Keywords/Search Tags:Brain-computer interaction, Brain-computer interface, Motor imagery, Hilbert-Huang Transform (HHT), Electroencephalogram (EEG)
PDF Full Text Request
Related items