EMG- blased Fatigue Adaptation in Admittance Control of Hand Rehabilitation

Prolonged muscle activity affects the neuromuscular system's ability to produce maximum force and will cause fatigue in the muscle. Rehabilitation exercises must contain certain repetitive hand movements with high intensity. Therefore, rehabilitation therapy with constant task difficulty levels may cause damage to the post-stroke patient. Thus, the problem is to design a system that is adaptable based on the user's condition. Using electromyography (EMG) signals to make better communication with rehabilitation robots has been long established. A modifying controller, according to these signals ensures a proper and safe exercise for the operator. An admittance controller with an adapting strategy utilizing a machine learning algorithm on a 2DOF robot is presented. Twenty healthy people participated in the experiment, and it is shown that the controller can provide adequate user assistance.