How to use neural networks to study the reliability of dynamic systems
Résumé
Presents an application based on neural networks. The goal is to determine the event sequences that induce the failure of an industrial system, and also to calculate the different parameters of a reliability analysis. Associated with a flow diagram and tested with an ISdF test case, it is shown that this model is able to give results comparable to the ones obtained by classical methods. After an introduction of the different methods used in the reliability domain and their limitations, the studied case is presented. This model is well adapted to the study of dynamic systems using Monte Carlo simulation. Then the structure of the model "flow diagram and neural networks" is shown. Finally before concluding, the results are compared to others obtained by several methods.