The research work focuses on the use of artificial neural networks in solving the inverse kinematics problem for the 3L flat robot manipulator on the Cartesian plane. The manipulator, designed by this method, will soon be equipped with an effector and will become an integral part of the agricultural robot complex.
In a neural network on a Cartesian plane, the coordinates and direction of the point in question are used as input data, the network, in turn, calculates the corresponding angles between the links in order to bring the effector to the desired point. The kinematics of the manipulator’s movement is planned by a neural network.
This new method of controlling the manipulator using a neural network allows you to reduce the positioning error in non-standard conditions.
Keywords: inverse and direct problems of kinematics, manipulator, neural network, 3-link manipulator, effector.
