Performance analysis of a surface EMG based control scheme of an exoskeleton robot

Abstract

In the present research work, the main emphasis has been given to study different issues of SEMG controlled exoskeleton robots and analysis of SEMG signal for discriminating motions of arm/wrist acquired from different electrode locations on the arm. The work undertaken in the present thesis has been split into three parts; (i) the acquisition/processing of SEMG signal from the muscle of interest and interpreting the SEMG signal; (ii) Extraction of the feature vector; and iii) development of an improved algorithm to achieve better efficiency and accuracy in classification of human upper limb movements. The simulation analysis of the entire proposed study has been carried out using MATLAB/SIMULINK platform. The performance of the time-domain features and the time-frequency domain features have been analysed on various types of classifiers. newline