High-level controller for an arm rehabilitation robot - positioning algorithms with respect to EMG data

Abstract

This paper touches upon the issue of dexterous control of rehabilitation robots. There have been proposed a high-level control system and algorithms used for training patients. Emphasis was made especially on the gravitation compensation algorithm, because of the low velocity and acceleration of movement parameters during rehabilitation. The direction of movements was estimated based on the model of surface electromyographic (sEMG) signal, which renders muscle activity very well. The model of manipulator is stable, behaves in a predictable way and, according to the EMG interpretation block, satisfyingly deduces the intention of motion and controls the model.