Over the last few years, parallel robots have been under increasing developments from a theoretical view point as well as for practical applications. They have good features such as high payload, high stiffness and high speed. However, a great disadvantage is their limited workspace. Singularities are abundant in parallel robots. When a mechanism is in a singular configuration, it loses stiffness completely and becomes uncontrollable. Singularities make the limited workspace of parallel manipulators even smaller. Redundant actuation has been proposed recently as an efficient approach to remove singularities over the workspace. It also has advantages of improving Cartesian stiffness, achieving uniform output forces, optimizing internal and external forces and so on....[ Read more ]

Over the last few years, parallel robots have been under increasing developments from a theoretical view point as well as for practical applications. They have good features such as high payload, high stiffness and high speed. However, a great disadvantage is their limited workspace. Singularities are abundant in parallel robots. When a mechanism is in a singular configuration, it loses stiffness completely and becomes uncontrollable. Singularities make the limited workspace of parallel manipulators even smaller. Redundant actuation has been proposed recently as an efficient approach to remove singularities over the workspace. It also has advantages of improving Cartesian stiffness, achieving uniform output forces, optimizing internal and external forces and so on.

Dynamic modeling and controlling are two fundamental issues of robotics. Dynamics of redundant parallel manipulators is much more complicated than that of serial robots due to multiple closed chains and redundant actuation. A new scheme to compute the inverse dynamics is proposed using Lagrange-D'Alembert's principle. The new strategy is more computationally efficient for real-time control. Independent joint control and model based control methods represented for redundant parallel manipulators are formulated, i.e. PD control, augmented PD control and computed torque control. The stability analysis of the proposed controllers is presented.

A two DOF planar parallel manipulator with redundant actuation was built as a test bed to investigate the dynamics and control of redundant closed-chain mechanisms. The dynamic model was formulated using the proposed scheme in the research. The unknown parameters of the system are obtained by both experimental measures and identification methods. To verify the estimated dynamic model and to evaluate the proposed control schemes, the control algorithms were implemented on the mechanism. The experimental results are analyzed.

There is little research about the dynamics and control of redundantly actuated parallel manipulators. The work done on the dynamic modeling and controlling will be valuable attempts in the field.