Identifiability Analysis of a Simplified Combustion Model*

Combustion is essential in modern society. It is the underlying process that powers much of our transportation and enables the creation of many goods we use on a regular basis. However, it is a difficult process to model. It involves many chemical species that are in turn involved in a complex network of chemical reactions. Models of this process include a large number of unknown parameters that need to be estimated from data. For example, we consider a “simplified” model involving nine species that has 231 unknown parameters. Because most of these parameters are irrelevant, we seek a simplified effective description of the combustion process. As a preliminary calculation we perform an identifiability analysis on this model. Our approach is to calculate the sensitivity of the model’s predictions to variations in the parameter values. I report on which types of parameters are identifiable in this model and the chemical insights this analysis provides.