|Titel:||Normal Vectors for the Robust Design of
The dynamic behaviour of chemical processes is usually investigated only after a candidate process design has been determined. The neglect of the dynamics in the design step is often compensated by oversizing equipment and using conservative design rules. This frequently leads to suboptimal process designs. In this work therefore a method is presented which integrates uncertainties, disturbances and the desired dynamic process behaviour directly in an optimization-based design step. The method is based on critical manifolds where the process behaviour changes qualitatively. The normal vectors of the critical manifolds are used to formulate constraints that ensure a minimal distance between a candidate operating point and the critical manifolds in the space of process and control parameters. The constraints guarantee that that the process remains in the domain of attraction of the optimal operating point and that state and output constraints are not violated despite disturbances. The method is successfully applied to various process models.
|Autor:||Dipl.-Ing. Johannes Gerhard|
|40 Abbildungen||15 Tabellen|
|Titel vergriffen, nicht mehr lieferbar|