The Conceptual Design Of An Ankle Rehabilitation Robot And Research On Its Control Strategy

The mechanical environment of ankle makes it easily hurt. In addition, surgery, stroke, spinal cord injury, cranial trauma and other neurological damage can also cause dyskinesia of ankle. The huge number of patients leads to the urgent call for rehabilitation. Currently, rehabilitation relies on therapists, which is not only tough work, but also insecure due to the lack of accuracy. Therefore, the demand for ankle rehabilitation robot is quite urgent.Firstly, this paper verified the necessity and effectiveness of rehabilitation based on anatomy and Neurorehabilitation, then proposed an ankle rehabilitation robot based on Stewart platform, enabling the rehabilitation of ankle for three rotational direction. Then, inverse kinematics of Stewart platform was analyzed. To get the exact position of this robot,forward kinematics was completed using Inv-Broyden method. Single rigid body dynamics was analyzed as well. Structural optimization was then done using genetic algorithm to determine structural parameters. Then model was established by Adams, and co-simulation with Simulink was completed in order to control the robot by simulation using hinge point control method. The hinge point control method worked out well with its bandwidth ranging to nearly 4Hz. Based on all above, human-machine force was further considered, and an innovative intelligent compliance control strategy was proposed, whose effectiveness in the ankle passive training was verified through cosimulation. The compliance control strategy can be widely applied in rehabilitation area, because it can achieve high accuracy of position when the torque between robot and human is much smaller than expectation, and achieve good compliance when the torque is big enough. In addition, the single neuron adaptive PID control strategy also achieved good results in the active training. That means it is possible to keep human-machine force at a given value when the patient is moving actively. Finally, the experimental performance of ankle rehabilitation robot was tested.