Computer experiments, that is, experiments performed using a mathematical model of a physical or chemical process implemented using a computer simulation are especially useful when extensive physical experimentation is either impossible or too expensive. In such cases one might perform limited physical experiments to calibrate the mathematical model and then perform extensive experiments using the simulation. Quantification of uncertainty for the outputs of such computer experiments is of great interest. The sources of uncertainty are in part experimental, measurement uncertainty in the inputs, and in part mathematical, that is, due to inadequacies of the underlying mathematical model. This paper presents methods for calibration and assessment of both types of uncertainty and demonstrates their use on two simple fire models.
Proceedings Title: Proceedings of the 2008 NCSL International Workshop and Symposium
Conference Dates: August 18-22, 2008
Conference Location: Orlando, FL
Pub Type: Conferences
uncertainty analysis, computer experiments, virtual measurements, calibration.