IDENTIFICATION OF DYNAMIC-MODELS FOR THE SHELL BENCHMARK PROBLEM

Dynamic process and disturbance models are identified for the 2x2 dist illation column described in the summary paper by Cott. Process models are identified using both classical least squares and the newer AUDI identification algorithm developed at the University of Alberta. The i dentified models are validated and gain-adjusted in the frequency doma in by comparing the magnitude spectrum of the models, \G(e(-i omega))\ , versus the spectrum obtained by fast Fourier transformation of the o riginal plant I/O time series. The noise/disturbance models are identi fied using standard time series tools and all models are verified by s howing that the filtered residuals (measured process output minus mode l based estimate of the output) are essentially white noise. The estim ated SISO models are all first-order and very close to the actual proc ess whether compared on the basis of step-response coefficients, model parameters or frequency domain plots. The control problem is so simpl e that two standard PID feedback loops with one-way (lead/lag) interac tion compensation produce results that are essentially identical to a 2X2 multivariable, model-based (GPC) controller.