Design And Control Of A Bionic Twisted String Driven Soft Glove

Rehabilitation glove can assist hand movement and restore the physiological function of the hand,which has important significance and broad application prospects in the field of medical rehabilitation.However,existing rehabilitation gloves have problems such as large volume and weight,poor fit and comfort when fixed to the hand,and complex and bulky driving parts,which undoubtedly increases the wearing burden of users.Therefore,this study was inspired by the skeletal structure and muscle motion mechanism of the human hand,proposes a twisted string driven soft glove,and conducts in-depth research on its design and control.The main research content is as follows:Starting from the physiological structure and movement mechanism of human hand bones and muscles,combined with hand rehabilitation theory,the design requirements of bionic soft glove were analyzed,and the main structure design and modeling of bionic soft glove were completed.A molding mold for bionic soft glove was designed and the main structure was completed through soft material molding.Research was conducted around its size,weight,range of motion,and other parameters.The results showed that the weight of bionic soft glove was only 85.1g,and the range of motion covered 85% of the human hand’s movement space,which can meet the needs of rehabilitation treatment.Inspired by the muscle tendon system of human hand,especially the driving tendon of deep digital muscle,which controls the movement mode of multiple fingers at the same time,the twisted string actuator was proposed as the driving method for twisted string driven soft glove,and the variable transmission ratio twisted string actuator and the multi terminal output twisted string actuator were designed.Through experiments,it has been proven that the variable transmission ratio twisted string actuator can achieve the transformation of output torque and output speed by changing the length of the string.The multi terminal output twisted string actuator has the function of adapting to the changes in load at both ends and changing the displacement size.Aiming at the problem of large error between the theoretical model and the actual motion data of the twisted string actuator,the invertible neural networks model was proposed as the control model of the twisted string actuator.The control model was used to complete the joint angle trajectory tracking experiment.The results showed that the average absolute error of the three joint angles of the index finger were 1.95°,2.62° and1.59° respectively,which proved that the control model had good accuracy and tracking performance.The prototype of the bionic soft glove system was constructed,and the passive rehabilitation experiment of the bionic soft glove was completed.The results showed that the bionic soft glove could effectively complete repeated rehabilitation tasks according to the rehabilitation plan.Through active rehabilitation training experiments,it was proved that the variable transmission ratio twisted string actuator has the characteristics of variable transmission ratio,and can complete the change of finger motion characteristics without changing the motor motion state.It was proved that the multi terminal output twisted string actuator has adaptive characteristics,and it can automatically adjust the displacement of the end of the finger and the force of the fingertip to adapt to the shape of the object when grasping the object.