The object of this project is the design and synthesis of multivariable robust algorithms for the control of positioning systems. This will allow increasing the systems performance. To this end, both linear and nonlinear techniques will be worked on.

With regard to linear techniques, the design of weighting matrices for Hinfty controllers will be systematized. Besides, working on the development of hybrid Hinfty /QFT controllers is intended.

Nonlinear techniques are quite less explored. In this project working on both the design of nonlinear Hinfty controllers and the control of Hamiltonian systems is proposed. These techniques are based on the energy attenuation/dissipation of the error signals. This gives a great versatility due to the generality of this concept.

Specifically, working on the design of new functions which permit attaining nonconventional objectives, such that periodic orbits generations, among others, is intended. In addition, nonlinear controllers with partial feedback of the information will be developed. This kind of controllers is particularly useful if the system state is not accessible. Additional aims are the synthesis of H2 controllers with Hinfty constraints and the study of inverse-optimal control techniques.

To validate the developed controllers, the application of the results to three practical devices with high industrial importance is intended: a high-performance robot, an industrial robot guided by visual feedback and a radio-controlled helicopter. All of them are multivariable systems with a great presence of nonlinear effects. .