A Computational Inverse Technique for Uncertainty Quantification in an Encounter Condition Identification Problem
W. Zhang; X. Han; J. Liu; R. Chen

doi:10.3970/cmes.2012.086.385
Source CMES: Computer Modeling in Engineering & Sciences, Vol. 86, No. 5, pp. 385-408, 2012
Download Full length paper in PDF format. Size = 746,176 bytes
Keywords Inverse problems; Bayesian approach; Penetration; Encounter condition; Uncertainty quantification.
Abstract A novel inverse technique is presented for quantifying the uncertainty of the identified the results in an encounter condition identification problem. In this technique, the polynomial response surface method based on the structure-selection technique is first adopted to construct the approximation model of the projectile/target system, so as to reduce the computational cost. The Markov Chain Monte Carlo method is then used to identify the encounter condition parameters and their confidence intervals based on this cheap approximation model with Bayesian perspective. The results are demonstrated through the simulated test cases, which show the utility and efficiency of the proposed technique. Since the uncertainty propagation in this identification process is efficiently explored, this technique can give us a clear indication of the degree to which we can trust estimates of the resulting encounter conditions.
PDF download PDF