Hiller, Manfred; Fang, Shiqing; Mielczarek, Sonja; Verhoeven, Richard; Franitza, Daniel:
Design, analysis and realization of tendon-based parallel manipulators.
In: Mechanism and Machine Theory, Jg. 40 (2005), Heft 4, S. 429 - 445
2005Artikel/Aufsatz in Zeitschrift
Design, analysis and realization of tendon-based parallel manipulators.
Hiller, ManfredLSF; Fang, Shiqing; Mielczarek, Sonja; Verhoeven, Richard; Franitza, Daniel


Tendon-based Stewart platforms represent an innovative manipulator technology that allows the handling of heavy loads with high acceleration and low energy consumption. Regarding aspects like light-weight structure or fast dynamics, tendon-based platforms with a parallel kinematic structure - i.e. Stewart-Gough-platform - represent a promising design and contain a large potential of various research topics in mechanics and mechatronics. This paper is a report about the state of the research project 'Tendon-Based Stewart-Platforms in Theory and Application (Seilgetriebene Stewart-Plattformen in Theorie und Anwendung (SEGESTA))'. The aim of the paper is the analysis of this type of manipulator under the aspects workspace, forward kinematics and trajectory planning. A major problem is to find acceptable force distributions in the tendons. This can be solved by a nonlinear optimization algorithm. The region where solutions are found is called controllable workspace. Other aspects of workspace include singularities, stiffness and autocollisions. Forward kinematics of such systems can most appropriately be described by kinematic mapping. It is possible to build kinematically overconstrained systems with the same kinematical behaviour (in particular, the same singularities) as normal ones; in the spatial case, even an infinite number of legs can be added. Trajectory planning must face a number of constraints, including workspace limitations, actuator and tendon limits, as well as smooth motion constraints. Together with the equations of motion and the actuator dynamics, this defines a set of possible trajectories. Time-optimal trajectories can be found via optimization strategies like i.e. simulated annealing. A manipulator prototype has recently been built and is actually tested in connection with various technical applications at the Institute of Mechatronics and System Dynamics at the University of Duisburg-Essen.