Seite - 215 - in Proceedings - OAGM & ARW Joint Workshop 2016 on "Computer Vision and Robotics“

Component J inkgm2 m inkg ls inm Motor/active link -combination 4.22×10−4 − − Blackpassive link 3.97×10−4 5.41×10−2 1.034×10−1 Redpassive link 2.90×10−4 4.48×10−2 8.32×10−2 Graypassive link 3.61×10−4 5.04×10−2 9.64×10−2 Table1.Overviewof thedynamicparameters theprevious joint and theCoGofeach component. Acomparisonofboth formulations implies, that behaviorof the joint error,whilecontrolledwithanaugmentedPDcontroller inminimalcoordinates formulation, is at least ei,max= 0.014rad (see Fig. 3). Additionally, there is a discontinuity in the motor torques at t = 0.6s, while the selection of independent coordinates is changed. From the simulation point of view, the redundant coordinates formulation is a quite better choice (c.f. Fig. 4). There are no discontinuities in themotor torques and joint errors are quite smaller (ei,max = 0.005rad). M3 M2 M1 t ins 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 e3 e2 e1 t ins 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 −0.015 −0.01 −0.005 0 0.005 0.01 0.015 Figure3. SimulationresultswithanaugmentedPD-controller inminimalcoordinates formulation 5. Conclusion In thispaperaproposal for aexperimental testplatformwithmodular setuphasbeengiven.Because of theusual reuseof similarmodules inparallelmanipulators,amethod tomodelentire systemswith subsystemshas been demonstrated. Differentmodel formulations for designingamodel-based con- troller havebeengivenand simulation resultswith anaugmentedPDcontroller inboth formulations havebeenpresented. Shortly, ageometric anddynamiccalibrationmustbedoneand thus, the simulation resultshas tobe validatedwithexperimental results. Furthermore, in thenear futurea replacementof thesearmswith elasticones isproposed. 215