Integrating Qualitative Simulation for Numerical Data Fusion Methods

Work described in this paper is aimed at developing a monitoring and fault detection system for the process plant. In particular this work concentrates on how to combine qualitative reasoning techniques with conventional numerical sensor fusion techniques. The advantages and disadvantages of using pure numerical or pure qualitative technique alone arefirst demonstrated. A method of developing a combined semi-quantitative model is then introduced. A number of algorithms are designed to analyse semi-quantitative predictions information. Finally, the semi-quantitative model is used to perform the system monitoring and change detection. The application of the algorithms developed is demonstrated by a number of eramples throughout the paper, based on real data.