Abstract
Reliability assessment is one of the necessary and critical parts in structural design under uncertainties. The methods for structural reliability assessment aim at evaluating the probability of limit state by considering the fluctuation of acting loads, variation of structural component or system, and complexity of operating environment. Latin Hypercube sampling (LHS) method as advanced Monte Carlo simulation (MCS) has higher efficiency in sampling. It will be chosen and applied in this paper in order to obtain an effective database for building Kriging surrogate models. In this paper, we propose an effective method to have reliability assessment by Latin Hypercube sampling based Kriging surrogate models. This method keeps the certain level of accuracy in prediction of the response of a structural finite element model or other explicit mathematical functions.
Highlights
Uncertainty is an inevitable issue in the process of manufacture, infrastructure, and engineering design
We find that the number of sampling in Monte Carlo simulation (MCS) or Latin Hypercube sampling (LHS) which is the original database for Kriging models is the most important factor to time cost
LHS method is more effective than MCS method for providing comprehensive original database to Kriging models
Summary
Uncertainty is an inevitable issue in the process of manufacture, infrastructure, and engineering design. P. Beaurepaire [5] attempted performing importance sampling technique in reliability based optimization of structure. The kriging models are interpolation models based on the assumption that there is a spatial correction between the values of the function to be approximated [18] With these models a known or sampled value of the limit state function to be approximated is exactly predicted. This paper presents an effective method to have reliability assessment by Latin Hypercube sampling based Kriging surrogate models. Latin Hypercube Sampling is utilized to build a reliable database for approximating the response of a structural finite element model or other explicit mathematical functions. Once the database is defined, we compare the relative performance of approximation methods to fit the probabilistic response of original models: namely, response surface method (first order and second order polynomial regressions) and Kriging models. Reliability assessment is predicted by the surrogate models, which heavily reduced computational cost and kept the certain level of accuracy
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Science Journal of Applied Mathematics and Statistics
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
