Control of the multifunctional prosthetic hand simulator via EMG signals

People may lose their limb because of injuries, accidents, medical conditions or congenital hereditary disorders. In recent years Researcher are carrying out many studies about design and control of prosthesis devices which take the place of the missing limb. Functional ability of prosthesis hand which mimicking biological hand functions increases depending on the number of independent fingers movements. With this perspective, in this thesis study six different finger movements were given to the prosthesis hand via bioelectrical signals and functionality of prosthesis hand was increased. Bioelectrical signals were recorded by Surface Electromyography method for four muscles with the help of the surface electrodes. The recorded bioelectrical signals were subjected to a series of preprocessing and feature extraction processes to calculate the maximum, effective, mean, variance and energy values of the EMG signals. In order to create effective cognitive interaction network between human and prosthesis hand, Classification algorithms have been developed which are Forward ANN, Progressive ANN, Radial based ANN, Exact Radial based NN, Probabilistic based ANN, Regression based ANN, Nearest Neighbor and Support Vector Machine classification and their performance compared with each other. A five-fingered and fifteen-joints prosthetic hand prototype has been produced via a 3D printer. And also a prosthetic hand simulator designed in the SimMechanics Motion control of both the simulator and prototype hand were realized in real time with about 70% success with obtained hand pattern information by classification via bioelectrical signals. Position control of motors connected to each joint of the Prosthetics hand is provided with designed PID and Sliding Mode Controller. Thus, an effective cognitive communication network has established between the user person and the real time pattern control of the prosthesis is provided by bioelectrical signals. Key words: EMG, Multifunctional Prosthetic Hand, SimMechanics, Cognitive Control