Design And Implementation Of Finger Rehabilitation Training System Based On The Motion Perception

The finger rehabilitation training system is a product of the intersection of rehabilitation,control engineering,robotics and other disciplines.It can effectively shorten the patient's rehabilitation cycle of finger function,reduce the labor intensity of rehabilitation doctors and improve training efficiency.It is a cutting-edge technology to solve the problem of finger rehabilitation training.It has great a value of researching and application.In the whole finger movement,the thumb function accounts for more than 50%,and the digital opposition function accounts for more than 80%of the finger function.The stroke patients could lost digital opposition function,part of tactile perception and proprioception function and loss of motion perception.It severely affected the daily life and work ability of patients.Aiming at the problem that the existing finger rehabilitation training system has a single motion perception parameter and lacks the digital opposition function,this thesis designs and implements a new finger rehabilitation training system basing on the motion perception.It focuses on solving a series of engineering and technical problems in the realization process of the system including mechanical structure design,finger force detection,active range of motion?AROM?measurement,circuit design,active and passive training mode design,touch and speech recognition human-computer interaction and 3D virtual hand model real-time follow-up.The main findings are as follows:?1?According to the analysis results of the physiological structure and movement characteristics of the fingers,I innovatively designed the mechanical structure which can achieve the training of digital opposition and the thumb.Comparing with other finger training devices,it has the advantages of simple structure and high safety.System test results show that the mechanical structure meets the design requirements of safety,reliability and comfort.?2?Aiming at the problem of sensory feedback loss in patients'finger movements,I have designed and implemented a finger motion sensing system based multi-sensor to realize the perception of finger angle,speed,position and finger force.It provides objective data for finger rehabilitation evaluation and rehabilitation training.Aiming the filtering problem of the finger force signal,I have proposed an improved limiting debounce averaging filtering method.The method combines the advantages of the limiting filtering method,the debounce filtering method and the average filtering method.It can effectively suppress the jitter and interference of the finger force signal,improve the accuracy of the finger force sensing system for the voltage signal acquisition.It can laid the foundation for the realization of active training algorithms.The experimental results show that the improved limiting debounce filtering method has stronger debounce ability and anti-interference ability comparing with the traditional limiting recursive average filtering method.the maximum error e_maxax of the sample signal can be further reduced by 31.8%,the variance S² can be further reduced by 37.4%.In addition,for the problems in finger AROM measurement,I have proposed a dual nine-axis sensor measurement methods of the finger AROM and have designed a new type of wearable measuring device.I have realized the measurement of the finger AROM and provided a new technical approach for rehabilitation evaluation of the finger function.?3?Basing on the finger motion sensing system,I have designed and implemented the finger rehabilitation training control system including hardware circuit,control system software,human-computer interaction system and host computer monitoring system.Traditional human-computer interaction relies heavily on PC,which makes the rehabilitation training system difficult to achieve householdization due to poor portability and high price.Therefore,this thesis designs a human-computer interaction system based on 7-inch intelligent touch screen and voice module,which realizes the householdization of finger rehabilitation training and rehabilitation evaluation.In addition,I have designed a 3D virtual hand model which is based on the virtual reality technology for presenting the details of the finger joint changes in real time.It can better feedback the information of the finger rehabilitation effect.The system test results show that the system software and hardware have reached the safety,stability and reliability of design requirements.The accuracy of voice command to control system is as high as 95%,90%and 84%in laboratory,hospital and residential respectively.?4?Under the guidance of the rehabilitation physician of the First Affiliated Hospital of Zhengzhou University,we have conducted a finger rehabilitation training experiment.The results of experimental show that the training device can drive the thumb and the index finger to complete the passive digital opposition training;when the finger force reaches the threshold,the motor rotates to assist the finger to complete the active training of the finger;when the thumb is trained,the joint movement angle is greater than the function.It further verifies the safety,stability and effectiveness of the finger rehabilitation training system under different training modes.