AK-P: An Active Learning Method Combing Kriging and Probability Density Function for Reliability Analysis

Abstract

An important challenge in structural reliability is to reduce the number of calls to the performance function. To reduce computational burden, surrogate models are commonly used. Kriging, one of the meta-models, is widely used as a surrogate for the original model in structural reliability analysis. In this paper, an active learning method combing Kriging and probability density function is proposed to improve the computational efficiency of AK-MCS. The proposed method, in general, provides a more efficient way by selecting the next best point effectively and adding it to the design of experiments to update the surrogate model more accurately. One example is used to demonstrate the efficiency of the proposed method.