Algorithm of Myoelectric Signals Processing for the Control of Prosthetic Robotic Hands

Abstract

The development of robotic hand prosthetic aims to give back people with disabilities, the ability to recover the functionality needed to manipulate the objects of their daily environment. The electrical signals sent by the brain through the nervous system are associated with the type of movement that the limbs must execute. Myoelectric sensors are non-intrusive devices that allow the capture of electrical signals from the peripheral nervous system. The relationship between the signals originated in the brain tending to generate an action and the myoelectric ones as a result of them, are weakly correlated. For this reason, it is necessary to study their interaction in order to develop the algorithms that allow recognizing orders and transform them into commands that activate the corresponding movements of the prosthesis. The present work shows the development of a prosthesis based on the design of an artificial hand Open Bionics to produce the movements, the MyoWare Muscle sensor for the capture of myoelectric signals (EMG) and the algorithm that allows to identify orders associated with three types of movement. Arduino Nano module performs the acquisition and control processes to meet the size and consumption requirements of this application.