The Research On Kinect-based Upper-limb Exoskeleton Rehabilitation Training And Daily Assistance System

There are many people around the world who cannot raise their arms by themselves,because of disease or accident.It will make the patients lose the function of ADL(Activity of daily livings)of upper-limb and caused a great burden on their lives.Based on the different characteristics of patients at different stages after the disease,it is necessary to provide them with a certain intensity of rehabilitation training and daily assistance.But in the current study,many researches focused only on the function of rehabilitation training,which has certain limitations to supply assistance to paitents.In view of the above problems,I had established an exoskeleton device which can supply self-rehabilitation training and daily assistance,based on Kinect 2.0 sensor.The specific work content is as follows:(1)The paper analyzed the model of human’s upper-limb,then determines the selection of the motor and the design of the exoskeleton device based on the parameters of upper-limb.Analyzes the kinematics of device and calculate the Jacobian matrix.(2)We built a human upper-limb exoskeleton device and analyzed the composition of its host computer and lower machine,then exerted the kinetics analysis to the device.To simplify the complex three-dimensional dynamic equations,the two-dimensional dynamic equations after gravity compensation has been used to replace it.After determining the angle-based impedance control method,the user can use the joystick to control the device to assist their motion.(3)We had selected the Kinect 2.0 sensor to set as the input signal terminal of the device to control the upper-limb exoskeleton device.By using the three camer mounted on the sensor,we can record the dynamic images with depth information,which can acquire the information of human and calculate the angle of human’s joint in real time.I had writhed the program of the sensor by using the Visual Studio,this software can control the sensor to obtain five angle values and record them.To get the more stable data of human’s upper-limb,the median filter is used to filter the acquired angle.Meanwhile,this paper also carried out four experiments based on Kinect 2.0 sensor,which verified the function of the software and the feasibility of the human-computer interaction system.(4)Besides,we transmit the angle value from Kinect to assistance device.Arduino and Processing IDE has been used in data transmission.Processing can be used to collect the data from the temporary files and spend to Arduino.We had calculated the relationship between the user and device and got the output angle value when the device is running.Based on this relationship,the paper builds three subsystems with Kinect 2.0 sensor: the self-rehabilitation training system;the daily assistance of the upper-limb exoskeleton device and the remote guidance system.By using these three systems,the patients can perform self-rehabilitation training and obtain daily life assistance for upper limbs.Through the usage of the upper-limb exoskeleton device,the patient’s quality of life can be improved,also the efficiency of treatment.