Search

This paper compares the performance of two different inferential schemes when applied to an experimental packed-bed reactor. The first scheme, proposed initially by Brosilow, is designed based on Kalman filter estimation. The second less traditional design uses an estimator computed from the Partial Least Squares regression method (PLS). The second approach was found to give superior performance when the nonlinear system under study is operated is a wide range of operating points. Due to the nonlinearity of the system it is essential to address the issue of robustness of the proposed schemes. This is formally done in this work using Structured Singular Value Theory. For the robustness analysis it is crucial to develop a realistic but not overly conservative uncertainty description. Since the PLS estimator uses a large number of measurements, a robust design based on the uncertainty associated with each one of the measurements would be very conservative. To overcome this problem a lumped uncertainty description is proposed which is identified directly from experiments.

An inferential robust control technique is applied to an experimental fixed bed reactor. The controlled variables are the exit concentration and the maximum bed temperature. Both controlled variables are inferred from one single temperature measurement. The location of this measurement is selected to optimize the performance of the closed-loop system when model uncertainty is allowed. Closed-loop experiments are conducted to test the robustness characteristics of the controller. From these experiments, the operating regions most sensitive to modelling uncertainty are determined. The nonlinear system characteristics can cause significant offset in the inferred controlled variables.