Control bioeléctrico de un robot para la asistencia de discapacidad de extremidades superiores.

Abstract

The aim of this research work was to design and implement an interface between acontrol system of a manipulator and treatment system of superficial electromyographicsignals (sEMG). To obtain data, signs sEMG were developed, which were taken fromthe biceps and triceps muscles through a Myo bracelet, the same that reads the muscularactivity by means of superficial sensors. The program that allows to process, identify,validate, and control the robot through signs coming from the arm, it is performed in theMatlab Simulink platform. In order to extract the characteristics of theelectromyographic signals an analysis was done starting from the time approaching andfrequency, thus, proving that the obtained parameters are: electromyographic integrated(IEMG Spanish acronym), absolute value average (MAV), quadratic average (RMS),median power frequency (MNP), average power frequency (MNF). All thoseparameters mentioned before have as objective to identify the several arm movements inextension and contractionpositions of the elbow. For classifying the movements, anartificial neuronal net was used in order to activate the freedom degrees of the robot. Amicrocontroller was employed for the interface between the system and the robot, whichsends information to the Matlab Simulink tool for the HIL and SIL tests. Also a videocamera was implemented for delivering data about the mouth and plate position. Thepositioning is transformed into meters for correcting the control actions and keeping anexpected path to feed the user. The accuracy in acquiring the electromyographic signalscan be better by placing an extra sensor at another part of the patient’s arm.