A SSVEP-Based BCI for Controlling a 4-DOF Robotic Manipulator

摘要

It is difficult to control high degree of freedom (DOF) robotic manipulator based on brain computer interface (BCI). This paper proposes a BCI system which utilizes steady-state visual evoked potential (SSVEP) for controlling a 4-DOF robotic manipulator. In order to elicit SSVEP responses, a stimulator panel circumferentially equipped with 12 LEDs as flickering sources is designed to provide visual stimuli. Applying canonical correlation analysis (CCA) method, elicited SSVEP responses can be decoded as movement commands. Furthermore, the novel commands mapping is designed for controlling end effector to move in 3D space in an efficient and accurate manner. Online experiments were carried out and three subjects participated in this study to perform move-grasp tasks in the simulation environment. Seven out of nine tasks were successfully completed within 50 trials with an average time of 156.71 seconds. The results show that it is feasible and efficient to control 4-DOF robotic manipulator with the proposed SSVEP-based BCI.