Linearized Inverse Dynamic Models for Robot Manipulators With Varying Link Parameters

Abstract

In this paper, two fast computational algorithms are developed for effective formulation for the linearized inverse dynamic robot models with varying (kinematic and dynamic) link parameters . The proposed algorithms can generate complete linearized inverse dynamic models for robot manipulators, taking variations (e.g., inexactness, inconstancy, or uncertainty) of the kinematic and dynamic link parameters into account. They can be applied to any robot manipulator with rotational and/or translational joints, and can be utilized. The computational complexity of these algorithms is only of order O(n) , where n is the number of degrees-of-freedom of the robot manipulator.