Monitoring steady-state gains in large process systems

BUSTOS, GERMÁN ANDRÉS; GODOY, JOSÉ LUIS; GONZÁLEZ, ALEJANDRO HERNÁN; MARCHETTI, JACINTO LUIS

Roma

Congreso; The 5th International Conference on Integrated Modeling and Analysis in Applied Control and Automation (IMAACA 2011); 2011

Stationary process gains are critical model parametersto determine targets in commercial MPC technologies.Consequently, important savings can be reached byacceding to an early prevention system capable ofindicating when the actual process moves away fromthe modeled dynamics, particularly when the actualprocess gains are no more represented by those includedin the model identified during commissioning stages. Asubspace identification method is used here todetermine the state-space model matrices that help todefine a gain matrix estimator. The pursued goal is amonitoring method capable to identify those gains of amultivariable model that start a mismatching condition,or that show tendencies to change already knownmismatches with the actual process. The anticipatedknowledge of these events should prevent processengineers about the eventual reliability of targetingoptimal process conditions with wrong gain estimations,and thus, help to localize the dynamic relationships forwhich an updating identification would be necessary.