Application of Advanced Estimation Techniques to a Chemical Plant Model

The paper is aimed at comparing some of the most promising and novel advanced techniques for estimation by assessing their effectiveness on the chemical process benchmark. Global and distributed implementations of the extended Kalman filter are the key elements of the work. In addition, the paper is also aimed at describing and developing a recursive implementation of the autocovariance least square algorithm for the on-line updating of the tuning knobs of the filter, demonstrating its relevance in the performance monitoring of chemical processes.