| Along with the development of Rehabilitation engineering,mechanism and electronics, researches on electromyographic artificial limbs go deep into intelligence. Surface electromyography (sEMG) is recorded from the surface of skeleton muscle by electrodes, it is the bioelectricity discharged by neuromuscular activities. It is a reflection of muscles functional state. Along with the development of computer, how to obtain the information of movement related from sEMG becoming more and more important. In traditional method, it was got from the characteristics of time domain or frequency domain which was based on supposes that sEMG is stationary signal, while it is not.In this thesis data acquisition platform of sEMG is tried, which contains hardware and software. The hardware contains electrodes, microcomputer, power and filters. Because of the characteristics of sEMG itself, common filters can't receive the perfect result, so special filter is designed based on VISHEEG16-24AMP. Restrained by hardware, software is written by C language to control the microcomputer, simply realize the acquisition of sEMG signal. As disturbance occurs during signal acquisition, transmission and reception, hardware can not meet our need ,digital filter is used to improve the capability of system before processing. We make use of filter module in SIMULINK toolbox carried by Matlab to realize.Wavelet theory is used to analyze the signal, which includes continuous wavelet transform (CWT) and discrete wavelet transform. Because of the redundancy of CWT which will save all the information of the signal, it is used by neural network as input. While discrete wavelet transform (CWT) is used to decompose the signal to different frequency according to time, it can describe different signals in both time domain and frequency domain. |
PDF Full Text Request