Inclusion of input saturation in the design of dynamically operable plants

Baker, Rhoda

Inclusion of input saturation in the design of dynamically operable plants

Master Thesis

2000

Publisher

University of Cape Town

Department

Department of Chemical Engineering

Faculty

Faculty of Engineering and the Built Environment

Abstract

Dynamic operability reflects the quality with which a plant can be controlled using feedback, and is a function of both the design of the plant and its associated control system. A plant designed on the basis of steady-state considerations alone could exhibit poor dynamic characteristics, leading to a loss of economic performance and a reduced capacity to effectively handle safety and environmental constraints. This motivates the need for the development of quantitative techniques for dynamic operability assessment, as well as its incorporation into procedures for process plant design. Optimization-based approaches to dynamic operability assessment permit simultaneous consideration of performance-limiting factors of nonminimum phase characteristics, input constraints and model uncertainty, and also provide considerable flexibility in the choice of performance criteria, decision variables and constraints. Recent work has incorporated operability requirements as constraints within a single optimal plant design problem formulation (Mohideen et at., 1997; Bahri et at.,1996). Young and Swartz (1997) considered the rigorous inclusion of input saturation effects in optimizing control. Actuator saturation introduces discontinuities in the system model and, to avoid potential problems using a sequential optimization approach, two alternative formulations were proposed for solving the problem within a simultaneous solution framework. Input saturation discontinuities were handled by the introduction of slack variables and their inclusion in either bilinear or mixed-integer constraints resulting in a nonlinear or mixed-integer linear programming problem respectively. The formulations were applied to a linear system with dead time to find the economically optimal operating point for a controller with fixed structure and tunings when disturbance deviations are taken into account. It was shown that using a strictly linear controller in this case would lead to an overly conservative estimate of the feasible operating range and consequently, a suboptimal operating point.

Description

Bibliography: leaves 109-113.

Keywords

Chemical Engineering

Reference:

Collections

Masters

Full item page