Development of an Upper Limb Exoskeleton System with Integrated Electromyography Signals

Abstract

Occupational accidents or injuries that result in prolonged immobility of the upper limbs can cause inconvenience in daily life. In addition to manual rehabilitation, exoskeleton systems can also be used to help restore motor function and correct related pains. This study proposed an upper limb exoskeleton system that utilizes surface electromyography (sEMG) signals generated by human muscles as the basis for controlling the exoskeleton. The use of the most direct signal from the wearer’s body allows for more accurate and intuitive control of the exoskeleton. The research primarily focuses on changes in the biceps and deltoid muscles and uses MATLAB to preprocess the signals, including filtering, differentiation, and calculating the root mean square (RMS) value. The support vector machine (SVM) classifier is then used to label and effectively distinguish movements. After experimentation, this method achieves an accuracy of approximately 82%, and it is found that when the system accurately identifies movements, it can assist the wearer in performing rehabilitation-related movements more effectively, improving muscle strength recovery speed and efficiency.