Electromyography based control of robotic arm using entropy and zero crossing rate

Abstract

With the increasing importance of robots in medical and industrial applications, the use of biological signals such as Electroencephalogram (EEG) and Electromyogram (EMG) for control of robots are being more sought for. In this paper, a novel method of robotic control using EMG signals is proposed. The user controls the robot using highly natural arm movements. This is done by attaching surface Electromyography (sEMG) electrodes to the user's arm and the signals recorded are processed and used for control. Zero Crossing Rate and Shannon Entropy were the features used for classification and a robot was successfully controlled using EMG signals from three different muscles of the upper arm. Also, a simulated human model was controlled using the proposed system utilizing the extracted D-H parameters.