Research on Power-Assisted Strategy and Device Based on Muscle Synergy

Surface EMG (sEMG) signals are non-invasive means of recording muscle activity that reflect the activation of human skeletal muscles. The purpose of this research is to study a dynamic assisted strategy based on muscle synergy, and to design an upper limb motion assisted device to achieve different assist tasks. Eleven healthy participants were recruited for the study. The participants were asked to perform grasping tasks at 9 target locations in the space and sEMG signals of the eight involved muscles were recorded. The non-negative matrix factorization (NMF) algorithm was applied to extract muscle coordination information during each corresponding experimental task. According to the muscle coordination information of all the participants' sEMG signals extracted by NMF decomposition, the corresponding upper limb motion tasks were decoded. The proposed method can decode the movement pattern of the human arm by considering the mapping relationship between the muscle coordination information and the joint motion, which may provide less effortful control of the robotic exoskeleton for rehabilitation training of individuals with neurological disorders or arm impairment.