Neural Modelling for Sensitivity Analysis in Reliability and Availability Calculations
The purpose of this paper is to present a multiparametric sensitivity analysis of the reliability and availability model of a properly selected system. This type of analysis gains importance for systems and processes whose behavior is strongly influenced by high-order interactions of the parameters. The approach followed consists in training a neural network on the results of a Monte Carlo reliability/availability simulation model and then exploiting the trained network to quickly provide, with reasonable approximation, the values of the reliability and availability quantities of interest for the sensitivity analyses. The work was a joint effort between the Department of Nuclear Engineering of the Polytechnic of Milan, Italy, and the Institute for Systems, Informatics and Safety (ISIS), Nuclear Safety Unit of the Joint Research Centre in Ispra (JRC), Italy which sponsored the project.
Bibliographic Reference: An oral report given at: European Safety and Reliability International Conference. Organised by: ESRA, Polytecnic of Torino. Held in: Torino, Italy, 16=-20 September, 2001
Record Number: 200013415 / Last updated on: 2001-06-27
Original language: en
Available languages: en