Optimal Bayesian Experimental Design for Combustion Kinetics

Experimental diagnostics play an essential role in the development and refinement of\nchemical kinetic models, whether for the combustion of common complex hydrocarbons\nor of emerging alternative fuels. Questions of experimental design—e.g., which variables\nor species to interrogate, at what resolution and under what conditions—are extremely\nimportant in this context, particularly when experimental resources are limited. This paper\nattempts to answer such questions in a rigorous and systematic way. We propose a Bayesian\nframework for optimal experimental design with nonlinear simulation-based models. While\nthe framework is broadly applicable, we use it to infer rate parameters in a combustion\nsystem with detailed kinetics. The framework introduces a utility function that reflects the\nexpected information gain from a particular experiment. Straightforward evaluation (and\nmaximization) of this utility function requires Monte Carlo sampling, which is infeasible\nwith computationally intensive models. Instead, we construct a polynomial surrogate for\nthe dependence of experimental observables on model parameters and design conditions,\nwith the help of dimension-adaptive sparse quadrature. Results demonstrate the efficiency\nand accuracy of the surrogate, as well as the considerable effectiveness of the experimental\ndesign framework in choosing informative experimental conditions.