Parallel robots have recently arisen as an interesting robotic architecture capable of performing task at high speed and precision. In order to exploit all the potential of these mechanism, model based control approaches are required. However, due to their complex structure, kinematic and dynamic modelling is a complex task, which usually leads to models with parameter uncertainties. In order to reduce the effect of parameter uncertainties, in this paper a redundant dynamic model based Extended CTC approach is proposed, and its stability and sensitivity analyzed for the 3RRR parallel robot. Results show that this approach provides more robustness than classical CTC approach to model parameter uncertainties.

A novel control architecture for parallel robots is introduced to fully exploit the advantages of these robots on high-speed and precision operation. A closed form of the dynamic model of parallel robots is difficult to obtain, due to the complex kinematic relations of these kind of mechanism. However, with the use of the extra data provided by sensors placed in strategic passive joints, kinematic and dynamic modelling can be simplified. The dynamic model can be used to implement advanced control techniques to improve the efficiency of parallel robots. In this paper, monoarticular and multiarticular control techniques are implemented on a 5R parallel robot, showing that the use of extra sensor data leads to a better and accurate control.

The authors present a general method for the automated singularity analysis of any mechanism at a given configuration. The procedure uses a base of the motion space. This is obtained from a velocity equation characterized by a geometric matrix. This special form of Jacobian matrix has some advantages for automatic implementation. This approach provides the degree of freedom associated with the singularity, uncontrolled motion, and kinematic dependencies. It also facilitates the choice of actuators and redundant devices. The method has been implemented in a computer program for kinematic analysis.