Research And Design On Control System Of An Lower Limb Rehabilitation Robot

With the improvement of rehabilitation medical needs, the research of rehabilitation robot has attracted more attention. Lower-limb exoskeleton rehabilitation robot can provide rehabilitation training for patients such as Hemiplegia, with movement disability of lower-limb. The lower-limb exoskeleton rehabilitation robot is a wearable robot, the robot provide rehabilitation training for patients who wear it by the robot control system. This paper focused on the design of the robot control system, built control system hardware and software platform, and designed control algorithms applied to different control strategies for the rehabilitation robot.The robot control system is in distributed architecture, consists of main controller and PC, node controllers, the underlying drives, motors and sensors. Node controller is located in each joint of the robot, used for the control of one single robot joint. This structure reflects the idea of the distributed control and improves the reliability and stability of the control system.For the design and analysis of the control algorithms for the robot control system, the dynamic model of rehabilitation robot was built based on the Lagrange Equation and the two-link model of the robot leg. The dynamic model describes the relationship between the torque and the position, velocity and acceleration information of the robot joint motion. The known and the unknown parameters were separated in the dynamic equation by linearization of the robot dynamic model.Trajectory tracking control algorithm was designed for the rehabilitation robot control system, including fuzzy PID control algorithm and adaptive control algorithm. The parameters of the fuzzy PID controller can be self-tuned, so the fuzzy PID controller has better immunity. Adaptive controller has better control effect for it can identify the unknown parameters of the robot model. Active and passive control strategies were designed for active rehabilitation training and passive rehabilitation training respectively. Finally, simulations and experiments were done to verify the effectiveness of the control algorithm.