Asymptotic Importance Sampling

by Marc A. Maes, Queen's Univ, Kingston, Canada,
Karl Breitung, Queen's Univ, Kingston, Canada,
Philippe Geyskens, Queen's Univ, Kingston, Canada,



Document Type: Proceeding Paper

Part of: Probabilistic Mechanics and Structural and Geotechnical Reliability

Abstract:

One of the main issues in reliability analysis is the accurate and efficient calculation of the multivariate reliability integral. The Asymptotic Importance Sampling Technique (AIS) selects an importance sampling density, based on theoretical consideration about the structure of the integrand of high reliability systems near its maximum points. The sampling is performed in the original variable space. Due to the asymptotic properties of the sampling density, AIS becomes increasingly efficient for smaller failure probabilities.

Subject Headings: Failure analysis | Probability | Integrals | Automatic identification systems | System reliability | Structural reliability | Integral equations

Services: Buy this book/Buy this article

 

Return to search