Investigation of the Possibility of Vector-Command Control Based on Forearm EMG

The work is devoted to solving the problems of increasing the depth and increasing the long-term stability of communication channels in human–machine interfaces, built on the electrical activity databases of the forearm muscles. For this, a method for analyzing electromyogram (EMG) signals is proposed, which combines vector and command control. The mathematical model for vector analysis of EMG is built in spherical coordinates based on the real spatial arrangement of the electrodes on the forearm, taking into account the possibility of a random phase shift during operation. Vector analysis of EMG is used to solve the calibration task of EMG sensors channels by the spatial arrangement of the electrodes and calculating the resultant vector of muscular forces. These forces are used as an additional information channel to set the movement direction of the control object operating point. Command control is based on gesture recognition by means of a pretrained artificial neural network (ANN) of the multilayer perceptron type. Processing results of actually recorded EMG signals by the proposed method are presented. There are given research results of correlation between processed signal fragments duration and the process of extracting the hand rotational movement information. It is proposed to use the signal duration of 250 ms. An algorithm of reassigning and calibrating the EMG channels amplification is proposed, which makes it possible to use further a once trained ANN for recognition and classification of gestures, while the position changing of electrodes between operation sessions is allowed. The work results can be used for calibration algorithms development, gesture recognition, and control of technical objects based on electromyographic human–machine interfaces.