EEG-Based Wheelchair Control Through a Brain- Computer Interface

A major challenge in two-class brain-computer interface (BCI) systems is the low bandwidth of the communication channel, especially this is important for controlling assistive devices such as a wheelchair or a mobile robot which requires multiple motion commands. The goal of this research is EEG-based wheelchair control through a SSVEP-based BCI. This brain control must provide high security and accuracy particularly for disabled people. SSVEP brain signal is windowed and for reducing noise and artifacts passed through a bandpass filter in preprocessing stage. Three methods including FFT algorithm, IT-CCA and an improved method based on combination of IT-CCA and Filter-bank are used for feature extraction. Computing Accuracy and ITR for three methods represents that improved method based on combination of IT-CCA and Filter-bank has the best performance. Finally the wheelchair and its physical environment are designed by Webots robotic simulation software and by sending and performing multiple commands, wheelchair motivation maneuver is completed.