One primary goal of structural reliability analysis aims at computing the probability of failure of a component or system with respect to some prescribed performance functions. In modern engineering, most performance functions usually resort to running an expensive-to-evaluate computational model, which leads to the infeasibility of simulation involving 103~6 runs. Surrogate models, such as response surfaces approach or Kriging are introduced as a substitute for the original model to cope with the computational cost. However, the analysis based only on the surrogate model fails to quantify the error made by this substitution. In addition, the existing variance technique in simulation is still less efficient in terms of complex computational model. Therefore, a hybrid algorithm combing Kriging surrogate model and importance sampling for structural reliability analysis is proposed to calculate the failure probability. The applications in several reliability problems proves the approach efficient, robust and accurate. Keywords-probability of failure; Kriging model; Importance Sampling; augmented failure probability; correction term
[1]
M. Eldred,et al.
Efficient Global Reliability Analysis for Nonlinear Implicit Performance Functions
,
2008
.
[2]
P. Bjerager.
Probability Integration by Directional Simulation
,
1988
.
[3]
R. Rackwitz,et al.
A benchmark study on importance sampling techniques in structural reliability
,
1993
.
[4]
J. Hammersley.
SIMULATION AND THE MONTE CARLO METHOD
,
1982
.
[5]
B. Sudret,et al.
Metamodel-based importance sampling for structural reliability analysis
,
2011,
1105.0562.
[6]
Sonja Kuhnt,et al.
Design and analysis of computer experiments
,
2010
.
[7]
Yan-Gang Zhao,et al.
A general procedure for first/second-order reliabilitymethod (FORM/SORM)
,
1999
.
[8]
V. Picheny.
Improving accuracy and compensating for uncertainty in surrogate modeling
,
2009
.
[9]
Nicolas Gayton,et al.
AK-MCS: An active learning reliability method combining Kriging and Monte Carlo Simulation
,
2011
.