Data Driven Approach for Inverse Kinematics in 2D and 3D

Abstract

Kinematics Analysis of Robotics is one of the most important tasks to compute. Inverse Kinematics, a field of Kinematic Analysis is considered to be difficult to obtain. The difficulty faced is due to the solution set generated by solving the Inverse Kinematics. Unlike Forward Kinematics, Inverse Kinematics provides a set of solutions and not a single solution. It provides a set of solutions as we can reach a place in space through many points. Despite having many methods to solve the problem, there is always been a doubt of which method outperforms all the other methods. It is seen that each system behaves differently while solving inverse kinematics. In this paper we provide insights on how systems with 7 Degrees of Freedom and 8 Degrees of Freedom behaves while solving the Inverse Kinematics in 2 Dimensional plane and 3 Dimensional space. The problem at hand is solved by using different methods that already exist and a comparison is done in the methods with the results obtained from the study.