Research Of The Brain Computer Interface Based On Alpha Wave

The Brain-Computer Interface (BCI) is a direct information communication and control channel established between human and computer or other electrical devices and it is a new communication system that does not depend on the brain’s normal output pathways of peripheral nerves and muscles. EEG-based BCI may provide an effective communication and control channels with world for the paralyzer, especially those "locked-in" but with intact ideation. BCI is winning more and more attentions.α wave (the frequency is between8and13Hz) will creat when we are closing eyes and obviously weaken or disappear when are opening eyes, so it is easily controlled. Ordinarily, the frequency of α wave is ten times bigger than other EEG signals, what is more, it is more seasonal, so it is more easily to observe its change. In this way, our research chooses a wave as analytic signal.Lab VIEW is a graphical programming language, it is widely industry, academia and research laboratories accepted as a standard data acquisition and instrument control.Under the environment of LabVIEW, we design the acquisition system and analysis system of α wave.Many noises are interfused into α wave when they are measuring. In order to remove the noises effectively, a nobel method, based on the ICA theory, is shown in the thesis. Methods:Power noise, ECG and EEG source signal are statistically independent, this features meet the requirements of ICA method, so we can transform the problem of denoising to be a problem of signal separation. Then under the environment of Matlab, use FastICA algorithm with a fast convergence to remove power noise and ECG artifact from a wave as an independent component. By the means of ICA method for noises separation, power noise and ECG artifact are basically been removed with no harm to the details of EEG signals.For a wave recognition, improved BP nerual network is used as classifier. We use the Matlab software to train the natural network. From experimental results we can get convergence curve of network and higher recognition rate. Experiments show that the use of BP network to classify a wave is entirely feasible.