Upper extremity arm prostheses's real time control using A.N.N. with a circuit based on DSP

ABSTRACT The aim in prosthetic device development is to build a fully functional reapplication of the lost limp. There have been various studies in this area. However, due to the recent developments in CPU speed and in the area of parallel processing has increased the study over myoelectric signals and has made possible the development of real time products. Especially, with the newly introduced DSP architectures this type of work can be done more easily. In this study a real time upper extremity prosthesis circuitry is developed with an Analog Device Inc. SHARC series DSP floating point arithmetic processor. As it is known, as muscle fatigue is getting increase an agglomeration takes place toward lower frequencies in the frequency spectrum of EMG signal. Therefore our study includes the analysis in a small time segment. Elbow extension, elbow flexion forearm supination, forearm pronation, operatings of the prosthetic arm is controlled trough the classified electromyogram signals. Keywords: Myoelectric, EMG, Prostheses, DSP, ANN.