A novel surrogate-model based active learning method for structural reliability analysis

Computational Analysis of Predicting Stress Intensity Factor Using Hdmr Technique

“…This demonstrates that the random sample points produced by the optimal design methods of the Latin hypercube can fill the design space uniformly. Reliability analysis has often employed the active learning method surrogate model because it offers a good trade-off between efficiency and accuracy (Hong et. al., 2022).…”

A novel learning function for adaptive surrogate-model-based reliability evaluation

“…In this study, we take into account the following two stopping criteria: The availability of computational resources is often constrained due to the expensive cost linked to evaluating LSFs, including the time-consuming finite-element analysis (FEA) and physical experiments. Consequently, to maximize the efficiency of resource utilization, the proposed method will cease execution upon reaching a predetermined maximum number of evaluations.Another stopping criterion was developed based on the probability of failure [48]. It mainly has two advantages.…”

“…Another stopping criterion was developed based on the probability of failure [48]. It mainly has two advantages.…”

“…When external loads are applied to the truss, the deflection at the red node depicted in figure 11 must not exceed a specified threshold of 0.11. The LSF is shown as follows: Gfalse(P1−P6,A1,E1,A2,E2false)=0.11−|delta(P1−P6,A1,E1,A2,E2)|.

Figure 11High-dimension truss structure [48].

…”

“…Consequently, to maximize the efficiency of resource utilization, the proposed method will cease execution upon reaching a predetermined maximum number of evaluations. (2) Another stopping criterion was developed based on the probability of failure[48]. It mainly has two advantages.…”