A general failure-pursuing sampling framework for surrogate-based reliability analysis. (March 2019)

• Record Type: Journal Article
• Authors: Jiang, Chen; Qiu, Haobo; Yang, Zan; Chen, Liming; Gao, Liang; Li, Peigen
• Is Part Of: Reliability engineering & system safety. Volume 183(2019)
• Page Start: 47
• Page End: 59
• Publication Date: 2019-03
• Subjects: Reliability analysis -- Surrogate model -- Failure-pursuing sampling framework -- Model-free response-distance function -- Design of experiment; Reliability (Engineering) -- Periodicals; System safety -- Periodicals; Industrial safety -- Periodicals; Fiabilité -- Périodiques; Sécurité des systèmes -- Périodiques; Sécurité du travail -- Périodiques; 620.00452
• DOI: 10.1016/j.ress.2018.11.002
• Languages: English
• ISSNs: 0951-8320

Abstract

Highlights: A general failure-pursuing sampling framework is proposed for efficient reliability analysis. The framework always pursues the local region sensitive to the prediction accuracy from a global perspective. Various active learning functions can be adopted to locally exploit the sensitive Voronoi cell. A model-free response-distance learning function is developed to avoid relying on the characteristics of different surrogates. Abstract: Design of experiment and active learning strategy are vital for the surrogate-based reliability analysis. However, the existing sampling and modeling methods usually ignore some useful information that can guide the choice of training samples, or heavily rely on the characteristics of surrogates. These lead to the inefficiency of sampling strategies or limit the application respectively. Therefore, this work proposes a failure-pursuing sampling framework, which is able to adopt various surrogate models or active learning strategies. In each iteration, it organically and sequentially takes into account the joint probability density function of random variables, the individual information at candidate points and the improvement of the accuracy of predicted failure probability. To measure the probability of the improvement, a global predicted failure probability error is proposed based on the real-time reliability analysis result. Furthermore, Voronoi diagram is applied to partition the sampling region into some local cells for keeping the uniformity of the training samples. Besides, a model-free response-distance function is developed and combined with the framework to avoid relying on the characteristics of surrogates, such as the statistical information provided by Kriging. Finally, four examples are investigated to demonstrate the applicability, stability and generality of the proposed method.