Dynamic optimization with a simultaneou method: Application to a heat exchanger

Abstract

In this work we use a numeric method based on a combined discretization and simultaneous dynamic optimization approach to solve a system consisting of partial differential and algebraic equations. The spatial derivatives are discretized by finite differences while the resulting DAE (Differential-Algebraic Equations) optimization problem is transformed into a large-scale NLP (Nonlinear Programming) problem through collocation over finite elements. This method is implemented in a computer package resident in a remote computer located at the Department of Chemical Engineering Carnegie Mellon University, which is accessed via a high-speed internet connection (Internet 2) from a client computer at the Centro Regional de Investigaciones Basicas y Aplicadas Bahia Blanca (CRIBABB). We have applied this strategy to the resolution of the dynamic optimization model of a gas - gas heat exchanger, which is part of a larger model under development. The goal is to minimize the transient between two set points of an outlet stream temperature. The dynamic model provides profiles of controlled and manipulated variables which are in agreement with available data, and the remote optimization system performed very well.