Abstract—Accuracy of results from mathematical model describing a physical phenomenon is complicated to infer due to several parameters that affect the model. With the ever increasing complexity of the models, the uncertainty in model development and parameter values are increased. For an analytical model having various input variables, only a few of the parametric values are known and the remaining values are assumed as the best case values. A quantitative value for each parameter in the analytical model, ranking in importance, is required to validate the model output. In this paper, the accuracy of an analytical model is estimated quantitatively using the uncertainty and sensitivity analysis. The developed methodology was applied and analyzed for two cases, a fluid flow equation and a heat transfer model. It is shown in this paper that the accuracy can be quantitatively predicted for an analytical model and the input parameters in their range can be effectively judged.
Index Terms—Model accuracy, parametric importance, sensitivity and uncertainty analysis, weight percentage.
A. Goyal is with the Department of Mechanical Engineering, Indian Institute of Technology, Delhi, New Delhi-110016, India (e-mail: me1080494@mech.iitd.ernet.in).
R. Srinivasan is with Nalco Technology Center, Pune-411028, India. (e-mail: sramanathan@nalco.com).
Cite: Anurag Goyal and R. Srinivasan, "Uncertainty and Sensitivity Analysis to Quantify the Accuracy of an Analytical Model," International Journal of Modeling and Optimization vol. 2, no. 6, pp. 648-652, 2012.
Copyright © 2008-2024. International Journal of Modeling and Optimization. All rights reserved.
E-mail: ijmo@iacsitp.com