Manipulation of a robot arm in 3D space by using EEG signals

摘要

This paper proposes an innovative robot arm control method of 3D space manipulation via using Electroencephalograph (EEG) signals. This system contains three parts: the Brain-Computer Interface (BCI), visual feedback, and robot control platform. First, the BCI collects and analyzes EEG signals from a subject and then classifies them into two different robot control commands. Thereafter, a binocular camera is employed to capture the objects and robot end-effector positions, which are then abstracted into symbols, e.g., rectangles, lines, etc, and then displayed on a screen. From the screen, the subject can get the position feedback information of the robot hand as well as that of the object in order to make decisions accordingly. Finally, the robot control platform decomposes the 3D coordinate system into a lower dimensional coordinate system, therefore the robot hand can be controlled to move to any specific positions corresponding to the subject's EEG commands through forward and inverse kinematics theory.