Abstract
Elastostatic calibration of robots is crucial for achieving high-accuracy positioning. Stiffness identification is an important step in elastostatic calibration wherein the stiffness parameters are identified. These stiffness parameters are then used to predict and correct the pose errors due to load applied on the robot’s end- effector/platform at any pose in the workspace. Hence, these parameters must be estimated accurately.For stiffness identification of robots, the relationship between the load applied at the end-effector/platform and its resultant deflection is first parameterized. This is accomplished using an appropriate stiffness param- eter model. Redundant deflection measurements are then performed with the robot subjected to a known load. These redundant deflection measurements are then used to estimate stiffness parameters by employing a least squares technique.Scaling of vectors containing measurements and parameters is crucial for accurate stiffness identification. This report focuses on this aspect. This report is organised as follows: section 1 presents the basic mathe- matical framework for stiffness identification of robots. Sections 2 and 3 discuss about the necessary scaling of vectors containing measurements and parameters.