An Application of Genetic Algorithms and Direct Search Methods to Crack Parameters Identification in Electromagnetic Non-destructive Testing

摘要

This paper presents some results from the inverse problem solution, based on the application of genetic algorithm (GA) and mesh adaptive search (MADS) technique. The investigation is directed to the identification of defects in metallic structures by means of electromagnetic Non-destructive Testing (NDT) methods. The proposed solution scheme uses rough identification procedure (RIP) and finite element method (FEM) as forward problem solver. The validity of the approach is tested with information acquired from numerical and physical experiments.