Research On A Soft Wearable Exoskeleton System For Upper Limb

The wearable soft exoskeleton system is a new direction in the field of exoskeleton research.With its high degree of freedom of movement,low equipment weight and comfortable to wear,it is gradually becoming the best for marching,rehabilitation and sports,etc.In this paper,the application of upper limb rehabilitation training for stroke patients is developed.The wearable upper limb soft exoskeleton system is developed from the perspectives of exoskeleton system design,EMG signal acquisition and feature extraction,and robot system impedance control.In order to meet the training requirements of upper limbs for stroke patients,the wearable upper lamb soft exoskeleton system designed in this paper can recognize the flexion/extension of the elbow joint,the internal/external rotation of the shoulder joint,and the internal/external extension of the shoulder joint.The exoskeleton system is mainly composed of a driving device and a soft suit.The driving device pulls the anchor point fixed on the soft cloth through the Bowden line,and then drives the limb movement.The wearable upper limb soft exoskeleton system uses the EMG signal to detect the body's intent.In order to establish the correspondence between the human movement EMG signal and the motion mode,the original EMG signal is firstly filtered by the fourthorder Butterworth bandpass to remove the high and low frequency interference and DC components in the signal.Then,the signal is subjected to a sixth-order DB wavelet transform,and the root mean square value and the singular value are respectively calculated for the obtained wavelet coefficients to form an EMG signal feature matrix.Finally,the feature matrix is sent to the convolutional neural network for training,and the network model is obtained.In order to control the assisted output of the wearable upper lamb soft exoskeleton system to the patient's affected limb,the system uses an impedance control strategy.The system detects the end position of the patient's limb and calculates the ideal output force at that position by a dynamic model that satisfies the impedance requirement.The feedback of the output force to the ideal output force is then achieved by the inner loop force-based controller.Finally,the effectiveness of the wearable upper limb soft exoskeleton system was tested by establishing a test platform to detect the consumption of VO2(oxygen inhaled)and VCO2(carbon dioxide exhaled)of the human body with and without the assistance of the exoskeleton system.