Design of observer based controllers

This project deal with design of observers in connection with feedback control. The concept of loop transfer recovery (LTR) has been invesitgated for the design of observers as well as design of fixed order LTR controllers.
In the usual LTR setting, design rules are developed based on sufficient conditions for recovery only. A consequence of this is that the design rules might not necessary point out the 'best' controller from an LTR point of view. Another drawback lies in the assumed controller architecture in previous approaches. Thust, it has not been investigaged if different choices of controller types would yield better general performance, or, conversely. which special performance properties are associated with different classes of compensators.
The LTR concept has, in this project, been applied on continuous-time, discrete-time and sampled date systems. Further, there has been focus on both design methods as well as on new observer/controller architectures for increasing the controller performance,
In the area of design methods, new methods has been derived out from standard induced norm based methods.
Two new architectures has been derived for increasing the specific performance conditions of the controller. The first architecture is the socalled PI observer which make it possible to increase the performance at low frequencies. This is quite relevant in connection with non minimum phase systems, where it is not possible to obtain good performance at low frequencies by using a standard observer architecture. The other new architecture is related with fixed order LTR controller design. A linear matrix inequality (LMI) design approach for fixed order controllers has been derived.