Research On Adaptive Control Of Lower Limb Rehabilitation Robot Based On Human Gait Characteristics

In recent years,our country has entered an era of aging population and the number of people with lower extremity mobility impairment due to stroke or spinal cord injury has been increasing.In order to restore mobility to these individuals,rehabilitation assistance for the lower extremities is necessary.Manual assistance from physical therapists is effective,but the number of physical therapists is limited.In this context,robotic rehabilitation is increasingly being used to assist patients with movement disorders due to its efficiency and effectiveness.In this paper,in order to provide safe and effective rehabilitation training to patients,we developed a kinematic model based on human walking motion data,combined with a rehabilitation robot with a closed horizontal chain of lower limbs,and validated and analyzed different rehabilitation training control strategies through simulation experiments.The main research components of this paper are as follows:(1)Analysis of human lower limb movement mechanism.Firstly,the neuroplasticity principle of lower limb rehabilitation is analysed,and the movement and muscle conditions of patients’ lower limbs at different rehabilitation stages are studied to provide a theoretical basis for the later control strategy research;afterwards,the structure and gait cycle characteristics of human lower limbs are introduced,and the equations of the movement trajectories of hip,knee and ankle joints and the movement angle curves are fitted as the reference trajectories of the control system by combining human lower limb movement data,and the Opensim system to calculate the joint moment changes during the lower limb motion.(2)Dynamics modelling of the lower limb rehabilitation robot.Four motion modes were designed based on the structure of the closed-chain horizontal lower limb rehabilitation robot.A sketch of the single-leg mechanical structure was established and the dynamics analysis was carried out using the Lagrangian method.Due to the different characteristics of the passive and active rehabilitation phases,two different kinetic models based on passive training and active training were established in the modelling process,taking into account the coupling between the human lower limb and the robot linkage.(3)Adaptive controller design based on different rehabilitation stages.A passive training control strategy is designed for the pre-rehabilitation stage.Firstly,PD feedback control based on the computational moment method is proposed,and on the basis of this algorithm,the total unknown function of the rehabilitation robot set is established,and an RBF neural network is used for its approximation estimation to achieve adaptive compensation control.Finally,the stability of the algorithm is proved using the Liapunov function,and the effectiveness of the algorithm is analysed through simulation comparison experiments;an active training control strategy is designed for the later stages of lower limb rehabilitation.The impedance control module is combined with position control to improve the flexibility of the system,and human-machine interaction force analysis is carried out to analyse the effect of different values of each impedance parameter on the control effect,and the effectiveness of the algorithm is proved through simulation experiments.(4)Construction of the experimental system for the lower limb rehabilitation robot.Design the single-leg physical prototype of the rehabilitation robot,carry out component selection and design,and build the experimental platform to verify the simulated rehabilitation through passive motion experiments.