Optimal Implementation of On-Line Optimization

Abstract

Results from a theoretical and numerical evaluation of on-line optimization algorithms were used to recommend the best way to conduct on-line optimization. This optimal procedure conducts combined gross error detection and data reconciliation to detect and rectify gross errors in plant data sampled from the distributed control system. The Tjoa-Biegler method (the contaminated Gaussian distribution) was used for gross errors in the range of 3σ 30σ or the robust method (Lorentzian distribution) for larger gross errors. This step generates a set of measurements containing only random errors which is used for simultaneous data reconciliation and parameter estimation using the least squares method. Updated parameters are used in the plant model for economic optimization that generates optimal set points for the distributed control system. Applying this procedure to a Monsanto sulfuric acid contact plant, a 3% increase in profit and a 10% reduction in SO2 emissions were projected over current operating condition which is consistent with other reported applications.

1 Figure or Table

Cite this paper

@inproceedings{Chen2001OptimalIO, title={Optimal Implementation of On-Line Optimization}, author={Xueyu E Chen and Ralph W. Pike and Thomas A. Hertwig and Jack R. Hopper}, year={2001} }