Preferencje help
Widoczny [Schowaj] Abstrakt
Liczba wyników
Powiadomienia systemowe
  • Sesja wygasła!

Znaleziono wyników: 2

Liczba wyników na stronie
first rewind previous Strona / 1 next fast forward last
Wyniki wyszukiwania
help Sortuj według:

help Ogranicz wyniki do:
first rewind previous Strona / 1 next fast forward last
EN
This work presents the BDU technique (Bounded Data Uncertainties) and the tuning of the linear quadratic regulator (LQR) via this technique, which considers models with bounded uncertainties. The BDU method is based on constrained game-type formulations, and allows the designer to explicitly incorporate a priori information about bounds on the sizes of the uncertainties into the problem statement. Thus, on the one hand, the uncertainty effect is not over-emphasized, avoiding an overly conservative design and, on the other hand, the uncertainty effect is not under-emphasized, avoiding an overly sensitive to errors design. A feature of this technique consists of its geometric interpretation. The structure of the paper is the following, in the first section, some problems about the least-squares method in the presence of uncertainty are introduced. The BDU technique is shown in the second section and the LQR controller in the third. After that a new guided way of tuning the LQR is offered, taking into account the uncertainties bounds via the BDU. The consequence of this method is that both recursive and algebraic Riccati equations are modified. Finally, some examples are shown and the main conclusions and future work are commented.
EN
The Generalized Predictive Controller (GPC) [1], [2] belongs to the general class of predictive controllers. The authors have proposed an alternative (although equivalent) formulation for the GPC in state space [7]. This formulation is based on a robust observer [5], and the poles selection is closely related to the controller robustness. An important feature of predictive controllers consists of their ability to take explicitly into account hard constraints in their formulation. However, their design must be accompanied by a guarantee of feasibility. There are some papers which deal with (his problem [4], [9], [8], [3], although all of them suppose that the state of the process can be measured on-line. However, in some cases, the design of the GPC proposed by the authors cannot measure online the process states since they are artificial states, that is to say, not related to physical magnitudes. The authors in paper [6] extend the results of [3] to the GPC in the case where all the states are online measurable. So the state estimation will be presented employing the same ideas of this previous work [6]. When the states have to be observed with the robust observer proposed, the authors show that there appears in the analysis a linear but time varying system perturbed with the error in the initial estimation of states. This initial error belongs to a known and bounded set. The main result states that if it is possible to find a collection of non-empty sets K_j that converge to the maximal robust control invariant set when j increases, the feasibility of GPC control law is guaranteed for all the sampling instants. Finally, this result is verified in a numerical example with a 2 states process.
first rewind previous Strona / 1 next fast forward last
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.