Inverse Kinematics and Trajectory Planning Analysis of a Robotic Manipulator

In this work, we pretended to show and compare three methodologies used to solve the inverse kinematics of a 3 DOF robotic manipulator. The approaches are the algebraic method through Matlab® solve function, Genetic Algorithms (GAs), Artificial Neural Networks (ANNs). Another aspect considered is the trajectory planning of the manipulator, which allows the user to control the desired movement in the joint space. We compare polynomials of third, fourth and fifth orders for the solution of the chosen coordinates. The results show that the ANN method presented best results due to its configuration to show only feasible joint values, as also do the GA. In the trajectory planning the analysis lead to the fifth-order polynomial, which showed the smoothest solution.