Research On Control Method Of Pneumatic Artificial Muscle(PAM)-actuated Ankle Rehabilitation Robot

Pneumatic artificial muscle(PAM) is a new type of flexible actuator which is actuated by air. PAM has the advantages of light weight, high power/weight ratio, good security, low price, cleanliness. The output characteristics of relationship between the PAM's driving force and the length are similar to human muscles which have good compliance. Due to these outstanding characteristics, PAM has a good application prospect in areas such as bionic robotics and medical rehabilitation. In this article, the ankle rehabilitation robot uses PAMs as the actuators, this will not only be able to help the ankle injury patients to achieve accurate and effective rehabilitation training, but also be able to ensure the patients' safety during the training. It has a good practical value.In this article, the PAM-actuated ankle rehabilitation robot is a strongly nonlinear complex system and some parameters of the system are time-varying. The robot has two degrees of freedom and there is a redundant actuator. The paper focuses on the position control method and the force control method on the basis of the robot experiment platform, and it can implement different kinds of motion mode of ankle rehabilitation training. The main research work included:(1) According to the working principle and features of the PAM, the mathematical model of PAM is studied. The ideal model is studied firstly, and then the experimental mathematical model is established using hysteresis model, which becomes the foundation of the robot control; the basic hardware composition and the working principle of the robot experiment platform is introduced; the control software system is developed; analysis of the robot kinematics.(2) Based on the kinematics of the robot, a position control method is proposed, which can make the robot move according to the presupposed ankle rehabilitation exercise trajectory. Fuzzy control strategy is used in the position controller. The self-regulating mechanism is added to enhance the adaptability of the controller. The experimental model of PAM is used for compensation which can overcome the lag of PAM's position control, improving the controller's accuracy.(3) Based on the demand of ankle rehabilitation, a force control method is proposed. Due to the existence of redundant actuation, each branch of the robot's actuating force need to be optimized, and then the distribution of the PAM's actuating force is appropriate; the force control uses a position-based impedance control method, so the robot is capable of outputting demanded rehabilitation torque by tracking the three PAMs' actuating forces.The position and force control method of the PAM-actuated ankle rehabilitation robot has achieved a good control performance, and this will provide a good technical support in the real ankle rehabilitation application.