Issues Related to Parameter Estimation in Model Accuracy Assessment

Model Accuracy Assessment (MAA) is an important part of the modern verification and validation process. This involves not only evaluation of a validation metric comparing experimental versus simulation system response quantities, but also the determination of the adequacy of the model for its intended use (see [1] for a detailed description). We describe here some issues related to the use of parameter estimation on MAA in the study of heat flow in a 2D metal plate. We consider seven parameter estimation techniques, and show that various factors such as length of sampling time, parameter estimation method, etc. impact the MAA. The ultimate goal is to improve MAA techniques in aircraft structural health care monitoring using Bayesian Computational Sensor Networks. c © 2013 The Authors. Published by Elsevier B.V. Selection and/or peer-review under responsibility of the Scientific Program Committee of the International conference on Computational Science (ICCE 2013).