Neural Network based Inverse Kinematics Solution for Trajectory Tracking of a Robotic Arm

Abstract

Planar two and three-link manipulators are often used in Robotics as testbeds for various algorithms or theories. In this paper, the case of a three-link planar manipulator is considered. For this type of robot a solution to the inverse kinematics problem, needed for generating desired trajectories in the Cartesian space (2D) is found by using a feed-forward neural network.