Integrated system optimization and parameter estimation using a general form of steady-state model

Abstract

Based on a general form of steady-state model, the integrated system optimization and parameter estimation problem is formulated and two algorithms arc derived for determining the optimal operating point of a process. A set of model qualifications, which are very mild, is defined to ensure that the real process optimum can be obtained by solving an equivalent model based optimization problem. Optimality and convergence conditions of the algorithms are analysed. The practical utility of the approach is demonstrated by a simulation of a continuous chemical reactor. It is observed that the correct optimum is quickly achieved in spite of using an incorrect model.