sEMG-based design of an upper limb exoskeleton rehabilitation training system
Objective In order to realize the rehabilitation training of patients with different degrees of upper limb injury,an upper limb exoskeleton rehabilitation training system based on surface electromyographic signal(sEMG)is designed.Methods The mechanical structure of the rehabilitation training system is mainly composed of the back control part,the variable stiffness driver and the adjustable upper limb support part.The control system includes EMG acquisition,filtering,feature extraction and action classification and recognition.We first collect electromyographic signals and extract their time-domain features;Then,principal component analysis(PCA)is used for dimensionality reduction,and K-means clustering algorithm is used for action pattern classification and recognition;Finally,the stiffness measurement experiment of variable stiffness actuator is carried out,and the simulation experiment is carried out to verify the classification effect.Results The rehabilitation training system can adjust the stiffness independently,and the overall recognition rate of action pattern is 89.74%.Conclusions The rehabilitation training system has a high recognition rate of movement patterns,which can better drive patients to complete rehabilitation training.
万方数据
维普
