Controle de um braço robótico através de eletromiografia

Abstract

Electromyography (EMG) can be defined as a detection, analysis and electrical muscle pattern recording technique used as a diagnostics tool and for movement control, using biomechanics and robotics’ concepts. This paper presents the control of a robotic arm utilizing EMG signals obtained from the brachial biceps. A study was developed in superficial electromyography technique and the project, simulation and development of a signal conditioning circuit, divided in two steps, amplification and filtering, can be found on its final stage on a printed circuit board. Furthermore, a robotic arm was designed with one degree of freedom and a tool, which is a claw. For the development of a robotic arm control system, was collected data from different people, authorized by the ethics committee registrant 30162814.5.0000.5547. These data were collected using a LABVIEWTM application through a NI DAQ data acquisition board from National Instruments and sent to a graphic generator application used for the signal analysis. After that, the data were analyzed and a control algorithm was developed for the robotic arm control, using the threshold analysis to detect the presence of muscular activity. All of this culminated in a LABVIEWTM application containing an acquisition stage, threshold adjustment, and the arm’s control through the data writing on serial port. The arm’s servomotors control was embedded using Arduino Mega platform utilizing a code that performs the serial data read and sends it to the servomotors used on the robotic arm. Lastly, the control system presented effective results on its operation, offering a great possibilities gamma for future work.