O'Reilly logo

Applied Process Control by Michael Mulholland

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

9Optimisation

9.1 Introduction

In Chapter 7 a need was identified for the optimal choice of control moves in model predictive control (MPC). In the case of the linear quadratic regulator (LQR) (Section 7.6.5), an optimal control policy was obtained explicitly by means of dynamic programming. Likewise, an objective function was minimised explicitly to obtain optimal future control moves in the unconstrained version of the dynamic matrix controller (DMC) (Section 7.8.2). In the more general context of optimal predictive control of hybrid systems, the focus was on preparation of the problem specification for a numerical optimisation solution. Yet other instances of explicit optimisation have been encountered in plant data reconciliation (Section 6.3), Kalman filtering (Section 6.4) and recursive least squares model identification (Section 6.5).

Nowadays, problems of optimisation occur in the above historical ways, and in a growing number of new instances in the field of advanced process control (APC). At the top of the plant control pyramid in Figure 1.5 are various optimisers that operate either in real-time closed-loop or in an open-loop ‘advisory’ fashion. These include traditional linear programs to match feedstock/product requirements (e.g. in oil refining), product blenders, event schedulers, supply and product chain optimisers, and such mundane items as how to cut paper machine output into required lengths and widths. In general, these applications are moving away from ...

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, interactive tutorials, and more.

Start Free Trial

No credit card required