Reliability Estimation Method for Frame Structures Based on Adaptive Importance Sampling Simulation

• Published in: Transactions of the Japan Society of Mechanical Engineers Series A Vol. 67; no. 663; pp. 1868 - 1875
• Main Authors: PARK, Young Tae, OKUDA, Shoya, YONEZAWA, Masaaki
• Format: Journal Article
• Language: Japanese
• Published: The Japan Society of Mechanical Engineers 2001
• Subjects: Adaptive Sampling; Failure Probability; Frame Structure; Importance Sampling; Probabilistic Method; Simulation Method; Structural Reliability
• ISSN: 0387-5008; 1884-8338
• DOI: 10.1299/kikaia.67.1868

This study describes a reliability estimation method for the frame structures based on an adaptive importance sampling simulation. Both Normal and Non Normal probability distributions are treated in the original basic random variable space. First introducing a preliminary sampling simulation process for estimating the failure probabilities, the sample means and variances of the respective failure domains, several dominant failure modes are selected. For the selected failure domains, the local importance sampling densities are determined so as to have the same parameters as the estimated sample means and variances respectively. Then a multi-modal importance sampling density is constructed as a composite of the weighted local importance sampling densities by weighting the failure probabilities of the selected failure domains. The importance sampling simulation is carried out to estimate the structural failure probability and the respective local importance sampling densities are updated in the simulation process. Numerical examples are presented to show the validity of the proposed method.