Structural reliability analysis in engineering applications is characterized by implicit performance functions that are expensive to evaluate. Monte Carlo Simulation with Kriging model has gained widely attention in recent years. These methods successfully reduce the computation costs and ensure the accuracy at the same time. This paper makes comparison of EGRA, AK-MCS, IBS criterion 2 and its deformation. The results present the problems that EGRA tend to search a local optimization of the next best training point and all of the stopping criterions perform badly. The performance of AK-MCS is medium in all cases. IBS criterion 2 and its deformation are not stable methods. However, IBS criterion 2 is relatively more efficient in low dimensional, highly nonlinear and complex cases, while the deformation is quite good at low dimensional and relatively simple cases.
A comparison of several proposed reliability analysis methods combining Kriging and Monte Carlo Simulation (IEEE CGNCC)
01.08.2016
1232172 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Reliability Analysis With Monte Carlo Simulation and Dependent Kriging Predictions
Online Contents | 2016
|A Monte Carlo Simulation of the Kriging Model
AIAA | 2004
|