The Research of Rehabilitation Robots BCI Technology Based on EEG

Brain-computer interface (BCI) is a new type of man-machine interfaces and one of the key technologies of rehabilitation robot. In this paper, the BCI system structure, working principle and the key technology are discussed. The dummy matrix based on event-related potential P300 is designed, used the stimulation by the character matrix sequence, subjects produced the evoked electroencephalograph (EEG). The evoked EEG is collected through placing electrodes on the scalp and then is amplified and filtered. The characteristics of signals are extracted by applying independent component analysis (ICA) principles and fast ICA algorithm. The simulation experiment shows the signals extracted are better approach the desired signals. With characteristics signals we can judge the type of moving or sensing of subjects. Without the muscle and peripheral nerve participate directly, the rehabilitation robots that based on this technology can help the brain to control the external environment. The BCI technology studied by this paper helps patients establish a new mode of communication, exchange, and rehabilitation therapy is provided.