Artificial neural network application for material evaluation by electromagnetic methods

The eddy current nondestructive testing of conductive materials is well known problem. For some eddy current transducers for flaw detection the mathematical models were constructed and the inverse problem (IP) were formulated In order to solve those problems the artificial neural network approach, relatively new, was adopted. The public domain software was used to teach and test the network. It was proved that, a standard feedforward multilayered neural network is sufficient to identify the parameters of different kind defects with a satisfying accuracy. Neural network approach could be extinguished to impedance tomography and to eddy current tomography. Several examples illustrated abilities of proposed approach will be presented in the paper. Two artificial neural network reconstruction methods for electrical impedance tomography have been presented in this paper. The problem under study concerns the reconstruction of the conductivity distribution inside the investigated area using the information collected from the boundary. The first approach consists in ANN learning using electrical potential vectors, which were obtained from numerical solution of the forward problems. The second method using a standard feedforward multilayered neural network applies the circuit representation for the finite element discretization. Using the quadrilateral finite element, the neural network structure for EIT problem has been proposed. The advantages and disadvantages of both methods with respect to the classical approach have been discussed in details.