ResearchResearch Projects
Leistungssteigerung von Parallelrobotern mittels parametervariabler Strukturen basierend auf kinematischer Redundanz

Improving the Performance of Parallel Manipulators using Reconfigurable Mechanisms based on Kinematic Redundancy

Team:  Dipl.-Ing. Jens Kotlarski
Year:  2010
Sponsors:  German Research Foundation (DFG)
Is Finished:  yes

The goal of this research project is to increase the singularity-free workspace of parallel kinematic structures through kinematic redundancy and simultaneously to improve their accuracy, stiffness, and dynamics. In addition to appropriate analyzing and evaluation methods, procedures for the optimal open- and closed-loop control of kinematically redundant mechanisms with a variable structures have to be developed and applied in practice.

Details

Classical robots consist of a serial kinematic structure. As a result every actuator has to move each of its following links and joints in addition to the working load. Therefore, these machines quickly reach a performance limit when standards regarding dynamics and rigidity are raised. Parallel structures characterized by closed kinematic chains hold a solution. The working load is reduced due to fixed actuators or actuators close to the base frame. Thereby, the robot’s dynamics are improved. Moreover, position and orientation errors are not accumulated which results in a higher accuracy. However, parallel robots suffer from the presence of singularities within their workspace. Furthermore, several performance indices, e.g. the achievable accuracy, highly depend on the endeffector's pose.

                         


Hence, there may be regions within the workspace where the endeffector is unable to operate or where the manipulator can not achieve the desired performance. As a result, the useable workspace and therefore the workspace to installation space ratio, which is already quite small, is further reduced. To compensate these drawbacks redundancy can be applied.

The characteristics of singularities and of the robot's performance such as accuracy, stiffness, and dynamics highly depend on the geometry of the mechanism. By adding at least one actuated joint to one kinematic chain kinematic redundancy is realized.
As a result, the inverse displacement problem has an infinite number of solutions. Hence, reconfigurations of the mechanism can be performed selectively in order to avoid singularities and to affect the achievable performance directly. As a result, the useable workspace and therefore the workspace to design space ratio increases significantly.

                         

The goal of this research project is to increase the singularity-free workspace of parallel kinematic structures through kinematic redundancy and simultaneously to improve their accuracy, stiffness, and dynamics. In addition to appropriate analyzing and evaluation methods, procedures for the optimal open- and closed-loop control of kinematically redundant mechanisms with a variable structures have to be developed and applied in practice.

Videos

Simulations/Animations

Basics (exemplarily for a planar 3RRR mehanism):

Exploitation of redundant actuators:

Kinematically redundant 3(P)RRR prototype

Exploitation of the redundant actuator

Optimal excitation trajectories (dynamics) - starting in slow motion (1/2):

Exemplarily chosen trajectory to calibrate the kinematics using passive sensor data:

Further demonstrations: