کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
307440 | 513362 | 2016 | 15 صفحه PDF | دانلود رایگان |
• The Line Sampling method is coupled with a metamodel of the performance function.
• Correction coefficient accounts for the metamodel approximation.
• Metamodel failure probability is updated with the correction coefficient.
• The estimate of failure probability is asymptotically unbiased.
This paper presents an approach for reliability analysis of engineering structures, referred to as Metamodel Line Sampling (MLS). The approach utilizes a metamodel of the performance function, within the framework of the Line Sampling method, to reduce computational demands associated with the reliability analysis of engineering structures. Given a metamodel of the performance function, the failure probability is estimated as a product of a metamodel-based failure probability and a correction coefficient. The correction coefficient accounts for the error in the metamodel estimate of failure probability introduced by the replacement of the performance function with a metamodel. Computational efficiency and accuracy of the MLS approach are evaluated with the Kriging metamodel on analytical reliability problems and a practical reliability problem of a monopile foundation for offshore wind turbine. The MLS approach demonstrated efficient performance in low to medium-dimensional reliability problems.
Journal: Structural Safety - Volume 60, May 2016, Pages 1–15