Teleoperation control of an exoskeleton robot using brain machine interface and visual compressive sensing

Abstract

In this paper, a brain-machine interface (BMI) based teleoperation system is designed for an exoskeleton robot by combination of vision compressive sensing and adaptive fuzzy control technique, such that the exoskeleton robot could perform manipulation task via mind control of a human operator. In this paper, the compressed images are used as feedback in a non-vector space to invoke SSVEP (Steady-State Visual Evoked Potentials) electroencephalography (EEG) signals. Local adaptive fuzzy controller has been designed to drive the robot following the intended trajectories extracted from the human operator's mind, in the presence of coupled dynamics and actuator constraints. Extensive experiments have been carried out to test the effectiveness of the proposed method.