Desenvolvimento de uma interface para acionamento de um braço robótico a partir de sinais de eletromiografia

Abstract

Electromyography is the visual representation of the electrical activity generated by the movement (voluntary or not) of the muscles. Therefore, electromyography is a tool very useful for treating patients undergoing physical therapy or even in studies regarding muscle activities. The objective of this work is to develop an analog interface capable of to collect the electromyography signals, after the collection, the signals are processed by an algorithm able to identify moments of muscle contraction and identifying these contractions, it is a trigger signal is sent to a robotic manipulator. For this, the signs of a volunteer's calf muscle. These signals were collected using sensors of non- invasive EMG, then these signals go through a pre-processing step through an instrumentation amplifier and a VCVS topology bandpass filter (voltage-controlled voltage-source). After this pre-processing, the EMG signal is analyzed by an algorithm that identifies muscle contraction by calculating the mean value of signal voltage. If this contraction is identified, a trigger signal is sent to control a stepper motor. The method developed in this project presents satisfactory results and that can be replicated, it was possible to collect the myoelectric signals and make the identification of the muscle contraction. However, the use of more accurate tools is recommended during the bench testing stage.