Simple empirical nonlinear model for temperature‐based high‐purity distillation columns

Abstract

A very simple empirical nonlinear low-order model has been proposed for temperature-based high-purity distillation columns. This model is in first-order form with the process gain and time constant as nonlinear functions depending on operating conditions (where process measurement is currently at) and two limiting high and low temperatures (bottom and top temperatures). A rigorous distillation column simulation will be used to represent the plant. Some plant dynamic test data will be used to estimate the model parameters. Prediction capability of the proposed model will be compared to some other commonly used nonlinear models (Logarithmic transformations; Nonlinear autoregressive model with exogenous input (NARX); Wong and Seborg's empirical nonlinear low-order model)