Multifunctional Prosthesis Control System Development Base On EMG Signal Decoding

Using neural-machine interface technology and information decoding technology, mixing together with bionic-robot control technology, a new generation prosthetic control system for the amputee is developed. This is an important direction of Rehabilitation Engineering. Surface electromyogram(sEMG) signal is very stable and easy-availably, for these feature, It is becoming a research-hotspot for using sEMG as control signal of Human-Computer Interaction and the multifunctional prosthesis control system.The sEMG were chosen as the source control signals of multifunctional prosthetic and its characteristics were studied, the corresponding amplifier and acquisition system of electromyogram(EMG) has been designed. Through this acquisition system the multi-channel signal can be collected. Then the signal collected would be decoded by system and the classifier for system was designed. Finally, the virtual reality platform system and the multi-functional prosthetic real-time control system were developed.The basic characteristic of sEMG signal has been analysed, according to the characteristics of the EMG signal, the corresponding preamplifier and main amplifier were designed. The preamplifier is placed on acquisition electrode for amplifying and filtering the signal in time. The main amplifier is aiming to further amplify and filter the signal.The sEMG signal acquisition system has been designed, through this system, multiple signals can be monitored in real time and stored in the hard disk offline or online.The sEMG signal collected was decoded and its features in time domain and frequency domain have been researched. Several classifiers based on time-domain characteristics were designed and compared by experiment, one classifier with comparative advantages was selected as the classifier of the system.The virtual-reality based training system for prosthetic control has been developed on Simulink platform. This system provides a relaxant and enjoyable training environment for prosthesis-users. It solves the problem that needs long training time for the amputees to learn to operate a myoelectric prosthesis and the heavy mental burden they suffered from the training. In addition, this platform also can be used to investigate the effect of various dynamic factors in practical application of a prosthesis system and performance of the prosthesis control system based on EMG signal decoded.The myoelectric hand on the market has been modified for the requirement of system and prosthesis-users. The corresponding control circuit and drive circuit were designed, then the real-time control system of multi-functional prosthetic were designed. Through this system, amputees can control myoelectric prostheses freely in real-time.