Isaac Horowitz (Univ. of California at Davis, USA) and Alfonso Baños (Univ. of Murcia, Spain)

Quantitative Feedback Theoy (QFT) has developed since the sixties around the work of I. Horowitz, being today one of the recognized techniques for designing practical control systems in many technological areas. Its most important properties are: (1) DESIGN TO SPECIFICATIONS: Formulation of (a) the ranges of the Plant parameter and disturbance uncertainties to be combatted by the feedback design, and (b) the performance specs. to be achieved despite these uncertainties (2) Rigorous, but systematic, relatively simple step by step design, mainly in the frequency domain, easily doable by ordinary, practical designers (3) Great emphasis on COST of FEEDBACK, especially on Loop bandwidths and Sensor noise effects, and their minimization. (4) DESIGN TRANSPARENCY: Early in the design procees and at each step, the principal trade-offs are highly visible. The designer can choose between them as he proceeds, such as bandwidth vs compensator complexity, competing sensors (in multiple-loop design), according to the special circumstances of his problem. The course will give an introduction to the fundamentals of QFT overall, and then concentrate on uncertain nonlinear/time-varying control systems.

The uncertain nonlinear/time-varying control Plant is mathematically rigorously converted to an EQUIVALENTuncertain LINEAR TIME-INVARIANT (LTI) PLANT. The solution (compensators) for the resulting LTI design problem is guaranteed to solve the original Nonlinear/TV problem. Thus, relatively simple LTI design procedure is used in most of the design. In a second QFT technique, especially suited to plant disturbance attenuation, the uncertain nonlinear/tv plant is converted into a combination of a very simple LTI plant and "equivalent disturbances". Fixed point theory provides the existence theorems for these techniques, but are not needed in the design process.