Robustness evaluation of the reliability of penstocks combining line sampling and neural networks

Antoine Ajenjo (1, 2) , Emmanuel Ardillon (1) , Vincent Chabridon (1) , Scott Cogan (2) , Emeline Sadoulet-Reboul (2)

1 EDF R&D - EDF R&D
2 FEMTO-ST - Franche-Comté Électronique Mécanique, Thermique et Optique - Sciences et Technologies (UMR 6174)

The objective of this work is to conduct robustness evaluations on the reliability assessment of penstocks using the info-gap framework. In order to improve the induced optimization searches, three original line sampling procedures are proposed in order to address the complex limit-state function on which the failure probability depends. The proposed algorithms are proven to be well suited for the search of the multiple roots involved in the line sampling technique. Then, a classification and a regression artificial neural network are combined for predicting the roots in order to reduce the computational time engendered by robustness evaluations.

Mots clés: structural reliability; line sampling; info-gap; neural networks; robustness

Domaines: Probabilités [math.PR]

Antoine Ajenjo, Emmanuel Ardillon, Vincent Chabridon, Scott Cogan, Emeline Sadoulet-Reboul. Robustness evaluation of the reliability of penstocks combining line sampling and neural networks. 2022. ⟨hal-03778836⟩