| With medical technology innovation,rehabilitation robots are widely used in rehabilitation nursing,prosthetic design and rehabilitation treatment.As a new type of intelligent actuator,pneumatic muscles was flexibility,high power-to-weight ratio and safety.It has gradually been favored as an actuator for exoskeleton rehabilitation robots.An elbow rehabilitation robot driven by pneumatic multifilament muscles is developed in this paper,the main work contents are as follow:Firstly,based on McKibben pneumatic muscle’s working principle,miniaturized the pneumatic muscles and designed thin pneumatic muscle.And then based on the biological muscle tissue and the changes in the diameter of thin pneumatic muscle,pneumatic multifilament muscles was prepared.Multifilament muscles’isometric experiment and isopiestic experiment were completed on the test platform.According to the experimental data,the Polynomial Fitting Model of multifilament muscles were established.Secondly,analyzing the human upper limb structure,joint motion mode and range of motion,the design principles of the exoskeleton rehabilitation robot were proposed,the parameters and driving modes of the exoskeleton mechanism were determined.Designed the three-dimensional prototype by SolidWorks,and then machined and installed the rehabilitation robot.Kinematic model of pneumatic multifilament muscles’contraction and joint angle was established,and the elbow joint model of the elbow exoskeleton rehabilitation robot system was further developed.Then,the passive training of rehabilitation robot,that is,the tracking control of the joint rotation angle,was studied.When the rehabilitation robot moved in case of no-load,a tracking differentiator,a nonlinear extended state observer and a linear error feedback controller were designed.Compensate the system uncertainties estimated by the real-time observer to the error feedback controller.Next,considering the weight in the case of rehabilitation training for different patients,designed a tracking differentiator,a linear extended state observer and a nonlinear error feedback controller.The real-time estimated external load changes and internal system uncertainties were compensated to the nonlinear error state feedback controller.In Matlab/Simulink,the simulation of the elbow joint position tracking control under different expected input signals was performed,which proved the effectiveness and feasibility of the algorithm,and provided theoretical guarantee for patients’ passive rehabilitation training.Finally,an experimental platform for elbow rehabilitation robots was set up.Noload passive training experiments was performed,evaluating the safety of rehabilitation robots through trajectory tracking accuracy.People with different height and weight as experimental objects,a passive training experiment with load was performed,verifying the rationality and comfort of the mechanical design;analyzing the trajectory control accuracy and errors,evaluating the robustness of the rehabilitation robot,and verifying the effectiveness and feasibility of the control algorithm.Analyzing the experimental results and then evaluating the rehabilitation effect of the robot. |
