A Lumped Model for Dynamic Behavior Prediction of a Hybrid Robot for Optical Polishing

Abstract

Combining screw theory with rigid/flexible body dynamics, this paper presents a lumped model for predicting lower-order dynamic behaviors of a hybrid robot for optical polishing. The model is built upon two assumptions: (1) the kinetic energy arising from elastic deformation of a limb-body within parallel mechanism is negligible if it also undergoes the rigid-body idle motion; and (2) the lower-order dynamics of the wrist is primarily dominated by the torsional compliances of two revolute joints. These assumptions lead to a 6-DOF dynamic model that enables the pose-varying dynamic behaviors to be predicted in an effective manner. The computational accuracy of the model is validated via simulations, illustrating the suitability to use the hybrid robot to perform optical polishing operations.