Stiffness Modeling of Planar Robotic Manipulators: Model Reduction and Identifiability of Parameters

Abstract

The paper deals with stiffness modeling of planar robotic manipulators and calibration of the model parameters. It was proved that the conventional stiffness models, which are widely used in practice, are parametrically redundant and only limited subset of link/joint compliance matrices elements can be estimated. To overcome this difficulty, several model reduction techniques were developed and corresponding parameter estimation methods were proposed. The validity and efficiency of these methods were evaluated via relevant simulation study.