Iterative Inverse Kinematics based on Screw Rotors and the Extended Kalman Filter

This paper presents a novel iterative algorithm for the solution of the inverse position kinematics for n-degrees-of-freedom kinematic chains with revolute joints using the Extended Kalman Filter. The algorithm implements an analytic gradient calculated using Screw Rotors in the Conformal Geometric Algebra framework, which is then employed as an update rule using the Extended Kalman Filter. Experimental results were carried out in simulation using the PUMA robot model, showing the proposed method's effectiveness in generating joint references to displace the end-effector to a desired position while improving the existing convergence time.