Sensitivity Analysis for the Inverse Problems of Electromagnetic Nondestructive Evaluation

Abstract

Sensitivity analysis of the model-based inverse problem associated to electromagnetic nondestructive evaluation is dealt with. Some uncertainty of the arrangement is inevitable present (imprecise host material parameters, sensor mispositioning, etc.), and this induces uncertainty on the reconstructed defect parameters. The aim of this work is to present a methodology for the ranking of the different sources of random error according to their contribution to the reconstruction uncertainty. To this end, state-of-art mathematical tools of sensitivity analysis are applied, including Sobol’ indices, and a polynomial chaos expansion surrogate model to reduce the computational burden of the method. A numerical example drawn from magnetic flux leakage nondestructive evaluation is presented to illustrate the proposed methodology.