TESTING OF PARALLEL ROBOTS

The advanced techniques of measurement statistic methods are presented in this work. Nowadays more and more producers use parallel kinematic structures as a manipulators, machining machines or measurement instruments. This structures main disadvantage, when compare to serial kinematic structures, is its limited workspace. The workspace is even smaller because of the existence of singularities. For this reason one of the tasks of our research is to maximize space for measurement requirements. Classic model of the cube within the working space for testing robots according to the ISO 9283 is not suitable for special workspaces which parallel kinematic structures robots are because it covers only a little part of workspace. This paper consider the workspace of Tricept which is a parallel kinematic structure type robot designed at Slovak university of technology in Bratislava. Relatively small dimensions of the workspace of Tricept is the main disadvantage as well as others parallel kinematic structures. This paper describes the other ways to choose measured positions used for testing this type of robot. This article also briefly describes the design of experiments of Tricept to determine the impact of various factors on one of important indicator of quality which is the position of end effector of robot. As a main design of experiment was chosen Central composite design. The presented paper gives certain overview on adapting the estimates of measurement uncertainty in position error compensation and analyse sources contributing to the overall difference between desired and real position. The measurement uncertainty is based on various influences, such as measurement environment, temperature during operation, used materials and others known or unknown factors. Positioning accuracy is one of the most crucial aspects which affects the quality of product, although it is often claimed that parallel kinematic structures are more accurate than serial one it is still highly demand increasing quality.