Rudy Agustriyanto - Multivariable Inferential Feed-Forward Control

Document created by Rudy Agustriyanto on Aug 22, 2014
Version 1Show Document
  • View in full screen mode

  Publication Details (including relevant citation   information):

  Ind. Eng. Chem. Res., Vol.42, No.18, 2003,   PP4186-4197


  Two multivariable inferential feed-forward control strategies   are  proposed in this paper. In the first strategy, the   effects of  disturbances on the primary process variables   are inferred from  uncontrolled secondary process variables   that are measured on-line. In  the second approach, the   effects of disturbances on the primary process  variables   are inferred from the manipulated variables for those    controlled secondary process variables that have fast dynamics.   The  proposed strategies are particularly useful in   situations where some  disturbances cannot be easily and   quickly measured. Robustness analysis  of the inferential   feed-forward controllers and the selection of  appropriate   secondary measurements are discussed. Structured singular    value analysis is used in assessing the robustness of the   inferential  feed-forward control systems. The performance   characteristics of the two  inferential feed-forward control   systems are demonstrated by  application to a simulated   methanol−water separation column. In the  first system, the   effects of disturbances in feed composition (and feed  rate)   are inferred from tray temperatures, whereas in the second   system,  the disturbance effects are inferred from inventory   manipulations.  Nonlinear dynamic simulation results   demonstrate the superior  performance of these strategies.   Robustness analysis shows that using  multiple tray   temperatures can improve the robustness of the inferential    feed-forward controller, and this conclusion is confirmed   by  simulation.

  Address (URL):