| People with a physical disability caused by a stroke or other illness can often recover their limb function through rehabilitation.Rehabilitation robots can replace doctors to help patients complete the rehabilitation actions prescribed by doctors in accordance with the set goals,and well-working rehabilitation robots require the support of a complete human-machine control system,so that they can accept real-time instructions from patients and correctly perform the rehabilitation movements expected by patients.In order to achieve the above control objectives,this paper will carry out the research on the identification of kinematic parameters of the human-machine system coupling by combining the kinematic parameter coupling characteristics of the humanmachine system of the terminal traction rehabilitation robot,and complete the development of the intelligent control system of the terminal traction lower limb rehabilitation robot on this basis.The main research contents of this paper are as follows:(1)Lower limb rehabilitation robot human-machine system coupling kinematic modeling and parameter identifying.The structure of the human-machine system is analyzed,the coupling motion topology model of the human-machine system is established,and the coupling kinematics model of the human-machine system is established on this basis.Aiming at the structural error in measuring the posture of human lower limbs by sensor feedback,an error constant is introduced,a kinematic parameter identification model of human-machine system with error constant pairs is established,and the effectiveness of the recognition model is verified by recursive least squares method and PSO algorithm.(2)The identify cation of human-machine system based on neural network method.Aiming at the problem that the kinematic model is difficult to accurately represent the human-machine system model,the BP neural network is used to identify the humanmachine system,the BP neural network is trained by the inverse model method to establish the mapping relationship between the training space and the joint space of the human-machine system,and finally the established neural network model is used to study the model-free control of the human-machine system,and finally the simulation experiment is designed to verify the model-free control effect.(3)Lower limb rehabilitation robot intelligent control system design.The framework of the overall control system of the lower limb rehabilitation robot is established,and the design principles of the control system software and hardware are proposed according to the functional requirements of each subsystem,and the hardware selection and software design are carried out accordingly,and finally the passive control and active control of the lower limb rehabilitation robot are designed.(4)Construction of experimental platform and algorithm verification.The lower limb rehabilitation robot control system was integrated,and an experimental platform for lower limb rehabilitation robot was established.Based on the experimental platform,the effectiveness of the designed rehabilitation robot control system is verified,and on this basis,the established parameter recognition algorithm and the system recognition method based on BP neural network are experimentally verified. |
