Critique of chemical process control theory

Convex formulations for optimal selection of controlled variables and measurements using Mixed Integer Quadratic Programming

“…Control structure selection deals with the selection of controlled variables (CVs/outputs) and manipulated variables (MVs/inputs), and the pairings or interconnections of these variables [1,2]. A comprehensive review of input/output selection methods was provided by [3].…”

“…More specifically, the objective is to find a linear measurement combination, c = Hy (1) such that control of these indirectly leads to acceptable operation with a small loss L(u, d) = J(u, d) − J opt (d), in spite of unknown disturbances, d, and measurement noise (error), n y . If the original optimization problem is constrained, then we assume that all optimally active constraints are kept constant (controlled) and we consider the lower-dimensional unconstrained subspace.…”

Experimental Evaluation of the Inferential System for Distillation Control

“…It is generally necessary to add some form of dynamic compensation (Foss, 1973) in order to form an acceptable control system. For small changes, the process can be modeled as a linear dynamic system about its steady-state permitting the design structure of the inferential control system to be presented in the Laplace domain.…”

Process-Insights-Based Control Structuring of an Integrated Distillation Pilot Plant

“…More generally, the issue of selecting controlled variables is the first subtask in the control structure design problem. 3,4 In partially related works, Agrawal et al 5 proposed new thermally coupled schemes from a steady-state analysis only, and Herná ndez et al 6 analyzed the controllability of thermally coupled distillation plants.…”