Розроблення модуля нейронної мережі для керування роботом-маніпулятором

Abstract

The work is devoted to testing the efficiency of the developed neural network module for the built-in robot manipulator. The possibilities of moving the manipulator in the Study Data environment were considered. The coefficients of the angle displacement of the kinematic pair of the manipulator arm for training the neural network are determined. During the experiment, the proposed module of the manipulator grip positions was evaluated with respect to the given endpoint of the grip position.