Eddy Current Probe Parameters Identification Using a Genetic Algorithm and Simultaneous Perturbation Stochastic Approximation

Sid Ahmed Chaiba (1) , Abdelghani Ayad (1) , Djamel Ziani (1) , Yann Le Bihan (2) , Martin Javier Garcia (3, 4)

1 ICEPS - Intelligent Control and Electrical Power Systems
2 GeePs - Laboratoire Génie électrique et électronique de Paris
3 ISEND - Ingeniería y Sistemas de Ensayos no Destructivos - Ingeniería y Sistemas de Ensayos no Destructivos
4 UI1 - University of Isabel I

This study tries to identify the coil parameters using numerical methods. The eddy current testing (ECT) is used for evaluation of a crack with the aid of numerical simulations by utilizing the identification of these parameters. In this study, a comparison of the performance of the GA and SPSA algorithms to identify the parameter values of the coil sensors are presented. So, the optimization probe geometry is introduced in the simulation with Three-dimensional finite element simulations (FLUX finite element code) were conducted to obtain eddy current signals resulting from a crack in a plate made of aluminium. The simulation results are compared with experimental measurements for the defect present in a plate.

Keywords: Eddy current testing; Genetic algorithm; Simultaneous perturbation stochastic approximation algorithm; Hole crack

Domains: Engineering Sciences [physics]

Sid Ahmed Chaiba, Abdelghani Ayad, Djamel Ziani, Yann Le Bihan, Martin Javier Garcia. Eddy Current Probe Parameters Identification Using a Genetic Algorithm and Simultaneous Perturbation Stochastic Approximation. Journal of Nondestructive Evaluation, 2018, 37 (3), pp.55. ⟨10.1007/s10921-018-0506-0⟩. ⟨hal-02404991⟩