Pattern recognition of electromyography applied to Exoskeleton Robot

Abstract

Exoskeleton Robot is a robotic-assisted human-machine system, which can provide power to assist the movement of people. This paper aims to find a method of EMG pattern recognition used in Exoskeleton Robot. To the beginning, the EMG power spectrum ratio (PSR) is calculated as the EMG eigenvalue. Then the minimum error-based Bayes decision rule is used to determine the movement intention of human, implemented in Matlab. The research result shows that the power spectrum density rate and the minimum error-based Bayesian decision theory can recognize the EMG on the movement intention of lower extremity exoskeleton with the advantages of easy reality and fast compute by Matlab.