Research On The Control System Of Lower-limb Exoskeleton Using Tendon-sheath Artificial Muscle

Light and soft man-machine interface structures are widely needed for new-designed comfortable walking assisted devices for wearable rehabilitation/training devices for the wounded people.Based on tendon-sheath-based artificial muscles,a wearable rehabilitation training control system is set up and control algorithms are designed to assist patients in rehabilitation training.Firstly,the structure of the lower exoskeleton driven by tendon-sheath-based artificial muscles is introduced,and the changing trend of the parameters of the Tendon-Sheath-Based artificial muscle drive model is calculated.The dynamic model of exoskeleton and upper extremity are deduced with Lagrange method.Based on Coulomb's viscous model,a system impedance model is proposed and parameters are identified.Secondly,a hardware and software platform are designed to control system.According to the system requirements,the hardware and software design of the control system is completed to realize the motion control of each joint.Thirdly,the clinical manifestations and training requirements of the post-injury patients in the initial stage of rehabilitation training are analyzed.A fuzzy PD control algorithm is designed.In order to further improve the control effect,a sliding mode control algorithm is designed,based on the man-machine coupled dynamics model and proved by the Lyapunov function.Then,the clinical manifestations and training requirements of post-injury patients in the mid-rehabilitation training are analyzed,an admittance control algorithm is designed for active training.Finally,the human movement intention is determined through the detection of plantar force and muscle force,and the superiority of the designed sliding mode controller is proved by comparison of a fuzzy PD control,independent PD control and a sliding mode control experiment,and then the exoskeleton is wore to test the stability of system.And an active training experiments is carried out to verify the effectiveness of the admittance control algorithm.