Research On A Hybrid-Powered Ankle Rehabilitation Robot And Training Strategies

The contradiction between the increasing demand of the rehabilitation training of ankle joints and the insufficient resources of the manual rehabilitation training promotes the rapid development of ankle robots,but ankle rehabilitation robots and its training strategies still have some limitations.Although traditional ankle robots driven by pneumatic muscles are suitable for ankle rehabilitation training,it is difficult to provide three degree-of-freedom(DOF)motion for the ankle rehabilitation training.Subjects with ankle disorders can not realize the motion intention in the rehabilitation training,so the effective of the training is poor.Therefore,it is necessary to develop an actuated-from-above ankle robot,which can provide three-DOF motion,and based on which comprehensive training strategies are developed.To solve the problem that traditional ankle robots driven by pneumatic muscles can hardly provide three-DOF rehabilitation training,a hybrid-powered ankle rehabilitation robot with three-DOF is developed in this paper.The kinematics and dynamics model of the robot is constructed to develop the software and hardware system.The relationship between the rotating angles of end effectors and the pulling force of pneumatic muscles,the relationship between the motion frequency and the trajectory tracking accuracy,and the reachability of the rotating extreme position are studied through theoretical analysis and experiments.The workspace of the ankle robot can be adjusted from six directions to improve the comfort of the human-robot interaction training.The feasibility and the safety are verified by theoretical analysis,simulation analysis and experiment.In order to solve the problem that subjects with weak control ability of ankle joints can not conduct continuously active training,this paper proposes a passive-active control method of trajectory adjustment and trajectory tracking separately,which is applied to three-DOF ankle training.The compliant of the robot are assured through determining the threshold by experiments.When the human-robot interaction force applied by subjects exceeds the threshold,a new expected trajectory will be generated in the rehabilitation training based on trajectory adjustment: modifying the phase to make the trajectory to be reversed;increasing the frequency according to the interaction force.If the human-robot interaction force is less than the threshold,the passive training mode will be maintained.In the rehabilitation training strategy based on trajectory tracking,an impedance controller is designed to control the trajectory and training mode according to the motion intention of subjectsFor the limitation that subjects with ankle disorders can not realize their motion intention in rehabilitation training,this paper proposes a training strategy of ankle joints based on Steady-State Visual Evoked Potential(SSVEP)and the virtual reality(VR)game environment.The VR game environment is constructed to provide visual feedback,and set up flickering sources with fixed frequencies to inspire SSVEP signals.The external information transmission loop based on the VR game environment is constructed to transmit information between the robot and the brain.FFT is applied to analyze SSVEP signal to judge the motion intention of subjects.Finally,the feasibility of this training strategy is verified by experiments.To solve the problem that it is difficult to recognize SSVEP signal inspired by distracted individuals,this paper proposes a method to combine trajectory orientation features and SSVEP recognition result to judge the motion intention of subjects.Trajectory orientation features are determined through judging the stretching state of ankle joints.The motion intention of subjects are judged according to the correlation between SSVEP recognition results and trajectory orientation features.The experimental results show that the success rate of motion intention recognition of subjects is effectively improved.