Optimization of continuous polymerization reactors: Start‐up and change of specification

Abstract

Optimal policies for the control of large disturbances (start up and change of specification problems) in continuous polymerization reactors are determined numerically and compared. Several algorithms are implemented for the solution of two‐point boundary value problems characterized by stiff systems and singular controls. The performance of different policies and solution methods is evaluated.

Optimal grade transitions in a gas phase polyethylene reactor

“…Dynamic systems modelled by nonlinear differentiallalgebraic equations (DAEs) arise in many aspects of chemical engineering. Optimization of these dynamic models is commonly used to solve kinetic parameter estimation problems (Biegler et al, 1986) and optimal control problems (Bryson and Ho, 1970;Ray, 1981;Farber and Laurence, 1986). Such optimization problems are more difficult to solve than steady state nonlinear programming (NLP) problems, because some of the equality constraints are given by differential equations.…”

Design of Rapid-Transition Multiproduct Processes

“…The general theory of this type of dynamic optimal control problem is given in Ray (1981). The method is developed more explicitly for this kind of situation in the paper by Farber and Laurence (1986) for a CSTR undergoing a change in operating state. Sargent and Sullivan (1979) have presented a more easily implemented suboptimal control system design comprising of off-line optimization and feasible on-line feedback for the case of a petroleum refinery changing its product profile.…”