| For amputees,the physical disability not only cause great inconvenience to their lives and work,but also bring great pain and suffering to their spirit and psychology.Myoelectric prosthesis hand that use sEMG signal as its control signal,is an important technology to help amputees recover motor function of upper limb partly,and has very important practical significance for improving the amputee's daily life and mental status.The main content of this paper is to develop a real-time control system based on sEMG for a new generation of myoelectric prosthetic hand which is being developed in our laboratory.The prosthetic hand is designed with eight degrees of freedom,and every finger is driven in underactuated manner.The thumb has two two degrees of freedom,and the rest of the four fingers have only one degrees of freedom,and wrist joint with two degrees of freedom.This paper designed a distributed control system that is fast and flexible is built using STM32F407&STM8S207,and use SVM as pattern recognition algorithm to recognize five kinds of gestures.The configuration of the finger position sensor and proximity sensor is important to enhance the perception of the prosthetic hand.The specific implementation process is as follows:In order to realize real-time recognition of hand gestures in the control system of prosthetic hand,it is necessary to finish the training of SVM classifier on PC.The process is as follows: using four separate electrodes and s EMG acquisition platform to acquire sEMG from certain muscles of forearm under five gesture,including open,close,pinching,holding a bottle and hook.Then use the collected data to complete the training and test of SVM classifier in the MATLAB,and get the support vector machine classifier model finally.Using off-line datas to test,the recognition rate of five kinds of gestures are more than 85%.According to the requirements and technical requirements of the myoelectric prosthetic hand,complete the design of hardware and software for the control system of myoelectric prosthetic hand.The hardware of the control system mainly includes the main control module,the finger control module,the sensor module and the power supply circuit.The main control module uses STM32F407 as the main control chip,and its tasks include: pattern recognition of five gestures based on sEMG,coordination the movement of five finger,reading and analyzing the information of the proximate sensor,driving shese motor of wrist and communication with PC.The finger control module uses STM8S207 chip as the main chip,is responsible for the DC servo motor drive control,the finger position information acquisition,motor current signal acquisition and motor encoder signal processing.The sensor module includes the position sensor and the proximity sensor,and the position sensor using the angle sensor of TLE5012 B,the proximity sensor using the environment brightness sensor of APDS9930 that integrated LED and infrared proximity detector.Power supply circuit includes 12 V to 5V circuit and 5V to 3.3V circuit.The software part includes the main control module program and finger control module,main control module program includes gesture recognition algorithm program,proximity sensor reading and writing program,communication program,finger control module includes current detection procedures,DC servo motor control program,position sensor to read and write procedures and so on.On the platform of Visual studio 2010,uses the MFC basic class library to develop the software for the debugging of the upper limb prosthesis,and realizes the real-time controlling of the prosthetic hand and displaying these information of the prosthetic hand.Finally,complete test of whole control system based on it is completed,including sensor test,DC servo motor control experiment,computer test,pattern recognition experiment and prosthetic hand control experiment and so on.According to the above test results,it shows that the control system designed in this paper meets the design requirements. |
PDF Full Text Request