Parallel robots have been a very active research field over the last 20 years. Compared to serial robots, parallel robots have essentially two well-known advantages such as a greater precision in positioning and an increased rigidity with respect to the relationship between size and work load limit. The main drawbacks of parallel robots are their small workspace and also specific problems related to control planning.

Industrial applications  demand the increasing of the robot speed without losing the accuracy of positioning at the end-effector of the robot. Thus, the implementation of advanced model-based control schemes is required. Moreover, any realistic simulation and optimization of motion, including their physical constraints, require an accurate dynamic model. The accuracy of the dynamic model depends largely on the certainty in which the dynamic parameters of the robot are identified, namely; mass, location of the center of gravity, inertia terms and friction parameters. Among the techniques that have been proposed for their determination, the identification of dynamic parameters through experimental methods is the one that has provided better results. These identification techniques have been applied widely to the determination of dynamic parameters of serial robots. Their application to parallel robots has been given few attention. This dissertation is part of a research topic on the identification of dynamic parameters that has been carried out in the Department of Mechanical Engineering of the Polytechnic University of Valencia over the last 10 years.

One of the most important aspects in the identification is the design of the trajectories from which the data is collected. The aim of this thesis is to design experiments with the purpose of increasing the certainty in which the parameters are identified. In this way, the design of the experiment based on multicriteria techniques is presented.

One of the most relevant contribution of this dissertation it is that not always the use of a complete and complex dynamic model can lead to a realistic identification of the dynamic parameters. That is, it has been found, on the basis of the experiments that were conducted, that it is preferable to consider simplified models, but with parameters determined with more significance, to define the dynamic behavior of the mechanical system.

In this dissertation a methodology is proposed for identification of dynamic parameters for parallel robots based on a set of significant parameters. The most important aspects of the proposed methodology can be summarized as: 1) Starting from a complete and complex dynamic model, to reduce the model by means of statistical considerations, 2) Imposing  physical feasibility condition on the parameters to be identified.

Experiments have been carried out on two types of parallel robots, specifically, fully parallel robots with three degrees of freedom; a 3-RPS robot configuration, which was available from a previous work, and a 3-PRS one. The latter was designed in the experimental framework of this thesis. The implementation of the proposed methodological strategy yielded to a model which was verified by solving the inverse dynamic problem and subsequent comparison of the generalized forces with the actual control actions. Moreover, the formulation of the direct dynamic problem has been treated by considering the identified dynamic parameter. In both cases, the system response showed a high degree of correlation with the actual performance.