An Identification Method of Physical Parameters for nR Serial Robots

The identification of physical parameters for manipulators is critical to the control precision of robots. Almost all the existing theories and methods are indirect solutions based on iteration techniques. Most of these algorithms are difficultly adopted in real engineering due to un-reachable or low precision measurement in Cartesian coordinate space of robots. An identification approach of physical parameters for nR serial robots is presented with its detailed process. The approach decreases the accumulated errors owing to the separately measure of each axis. At same time, the values of joint variables are accurately captured while physical parameters are gotten, and these values are significant to increase the precision of HOME position of robots. In addition, not only this approach can be used in D-H model but also it can be applied to modified D-H model without any changes of other steps of the process except the changes of Eq.1 to 7. Tested cased have identification performance of the proposed method satisfies engineering requirements of robot control.