Study Of A Prosthesis Control System Based On Speech Recognition And Electromyography

The number of disabled people is growing every day. However, the amputees take a large part. The major rehabilitation approach for these people is to use the prosthesis. It has a significant meaning to design a prosthesis for amputees which can be controlled independently and neatly.The limb residual muscles that can be used to control the prosthetic limb are limited, so could not provide enough electromyography (EMG) signals for control of a multifunctional prosthesis. To solve this issue, a prosthesis control system based on the fusion of speech recognition and EMG signal was proposed and developed in this study. This new control method could not only improve the usability of the existing prosthesis, but also expand the range of prosthetic controlling signal source and simplify the training process.To solve the issue of that amputees have lack of electromyography (EMG) signals for controlling of the multifunctional prosthesis, this study is carried out to expand the range of control signals. The amputee’s own voice is introduced to switch prosthesis joint or action mode, and a couple or one of EMG signals is used to control action mode in the joint which is selected by voice signal. Effectiveness of the control system is verified according to the control experiments. The experiments reveal that the average action completion rate of able-bodied persons is91.18%±4.87%and94.64%±4.26%; the amputees’action completion rate is an average of84.12±7.47%and98.41%±0.45%. The pilot results show that the system could classify the prosthesis controlling action modes efficiently, and control the prosthesis independently. The prosthesis control system based on the fusion of speech recognition and EMG signal is rational, reliable and practical.