On-Line Component Estimation and Analyser Calibration for a Complex Industrial Processing Plant using Kalman Filtering

The paper presents an on-line analyzer system based on a Kalman filtering technique. The system is implemented on a computer, and is tested against an analyzer simulator. A mathematical model for the analyzer is established. An extended Kalman filter is performing both component estimation and analyzer calibration. This will improve the accuracy of the analysis.