par Vanbergen, Jean Pierre 
;Hanus, Raymond ;Cohen, Isaac
Référence Ricerche di Automatica, 13, 2, page (298-306)
Publication Publié, 1982-10
;Hanus, Raymond ;Cohen, IsaacRéférence Ricerche di Automatica, 13, 2, page (298-306)
Publication Publié, 1982-10
Article révisé par les pairs
| Résumé : | A method based on the use of an Adaptive Control Module (ACM) made thanks to the knowledge of the process state description is presented. The outputs of this module are the actual process commands and the inputs are constituted by an auxiliary control signal set. Each auxiliary control signal is computed by a usual SISO controller (e. g. PI type). The ACM structure and parameters are determined in order to obtain the desired dynamics between the auxiliary control signals and the process outputs. When the process inputs come into saturation, it is possible - thanks to the ACM - to project these input saturations into the state variable space and to choose among the set points specifications which of them may be lost and which must be held. The method is easy to handle. It enables non-interactive and feed-forward control. It can be applied with great success to bilinear systems. The example of a mixer plant shows the advantages of the method. |
