EMG Signal Processing of Myo Armband Sensor for Prosthetic Hand Input using RMS and ANFIS

The population of people with disabilities in the world, including disability of hand function, has increased significantly from year to year. Developing a prosthetic hand that is reliable, ergonomic, based on stream data input and functions like a human hand is a challenge in current research. The development of knowledge and technology has now reached a stage that allows the use of body signals to operate artificial hand systems. One application of the development of knowledge and technology is the use of EMG signals from Myo Armband sensors as input signals from the prosthetic hand. Implementation of this research requires two stages of research, namely initial research and further research. The initial research phase focuses on the capture and processing of EMG signals while further research focuses on preparing prosthetic hands with all the mechanisms needed. This research is an initial study that begins with electromyography signal taking using Myo Armband Device mounted on human hands. The signal obtained is then extracted using the Root Mean Square (RMS) method and classified using the Adaptive Neuro-Fuzzy Inference System (ANFIS). The results of this initial study were in the form of EMG signal pattern recognition. Recognition of the EMG signal pattern from the Myo Armband sensor using the RMS and ANFIS methods for 7 hand signals including Rest, Power Grasp, Hook, Pinch Grip, Tripod, Thumb and Index in this study resulted in an accuracy rate of 98.09 percent. Referring to the accuracy value obtained shows that the results of this initial study can be used as input in the prosthetic hand control system which will be developed in subsequent studies.