Robustness of Metallurgical Process Observers Based on Mass and Energy Conservation Constraints

The robustness (or reliability) of a process state observation system is its ability to generate estimates despite failures occurring to some sensors. A reliability function and a robustness index are proposed for quantifying the reliability of linear observation systems in the face of sensor failures. This index is calculated from an analysis of the weak and strong degrees of robustness of the observer equations. The robustness index calculation method is illustrated for linear mass balance equations of a complex plant flow sheet. and for bi-linear mass conservation equations for a mineral separation plant.