Configuration Optimization And Motion Control Of A Reconfigurable Cable-Driven Parallel Robot For Rehabilitation Training

Reconfigurable cable-driven parallel robot is a special type of parallel robots that use flexible cables instead of rigid links.Because the lightweight cables can be easily wound on the drum,the robot has some outstanding advantages such as large space,high load,high speed and Reconfigure easily.In the field of medical rehabilitation,when using a reconfigurable cable-driven parallel robot for rehabilitation training,different requirements are often put forward for the robot due to some factors such as different rehabilitation scenarios and differences in patients to be recovered.The reconfigurable cable-driven parallel robot needs to be optimized for different scenarios,so that patients can get the best treatment.In this dissertation,according to different rehabilitation scenarios,different rehabilitation indicators are selected to optimize the design of the reconfigurable cable driven parallel robot.Then,we developed a controller which can effectively solve the model uncertainty problem in rehabilitation training.The main contents of this dissertation are as follows:(1)When the reconfigurable cable-driven parallel robot is used for upper limb rehabilitation training,the performance space of the RCDPR is defined according to the further analysis of cable tensions.Then,we use the performance space volume as the optimization index to optimize the configuration of the robot by using the simulated annealing algorithm.Based on the configuration optimization strategy,the robot can work with the optimal configuration in upper limb rehabilitation training.In the simulation experiment,we solved the optimal configuration of the reconfigurable cabledriven parallel robot in different situations,drew the corresponding performance space,and analyzed the influence of parameters on the optimal configuration and performance space.(2)For the application of reconfigurable cable-driven parallel robot in lower limb rehabilitation training,we conducted a compound optimal design on the robot’s training trajectory and configuration.On the basis of physiological knowledge,we first established a simplified skeletal muscle model of the flexion and extension of the lower limbs,and then established a dynamic model of the lower limbs and the robot.We choose the normalized muscle energy and the energy consumption of the robot as the optimization index to optimal the robot’s rehabilitation training trajectory and configuration.In the simulation experiment,we verified the effectiveness of the optimized design by comparing the optimized trajectory with the traditional trajectory,the optimal configuration with the common configuration.In the actual experiment,we designed a virtual muscle device to simulate the patient’s actual muscles.Based on the optimized results,we conducted experiments to verify the difference between the optimized trajectory and the traditional trajectory,proved that the optimized trajectory is excellent in activating muscle energy.(3)Considering the uncertainty of the dynamic parameters of the patient’s limbs in actual rehabilitation training,we designed a adaptive robust controller to improve the control accuracy.Firstly,we established the dynamic model of the reel and pulley,and the dynamic model of the patient’s limb.Then,we designed an adaptive robust controller.In simulation experiments,the controller demonstrated good performance when the model was subject to uncertainty.In subsequent actual experiments,the tracking performance of the controller for optimal training trajectories was verified,and the experimental results showed that the controller was able to meet the requirements of practical rehabilitation training.