• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>AIAA Non-Deterministic Approaches Conference;AIAA Sci'Tech forum >Reliability Analysis Considering Tail Dependence over Space and Time
【24h】

Reliability Analysis Considering Tail Dependence over Space and Time

机译：考虑尾随时间和空间的可靠性分析

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Tail dependence is important for reliability analysis since failure events are usually located at the tail of distributions. The common practice of multivariate ARMA model and Karhunen-Loeve (KL) expansion method for the modeling of multivariate stochastic loads cannot capture the tail dependence between loads. This paper presents a combination of Vine copula and autoregressive-moving average (ARMA) for load modeling in time-dependent reliability analysis, in problems with a vector of correlated non-Gaussian stochastic loads. Univariate ARMA models are used to maintain the correlation of stochastic loads over time. The vine-copula model can account for not only correlations between different univariate ARMA models, but also the tail dependence of different ARMA models. The proposed Vine-ARMA model is able to flexibly model a vector of high-dimensional correlated non-Gaussian stochastic processes with the consideration of tail dependence. In order to overcome the challenges in computational effort, a recently developed single-loop Kriging (SILK) surrogate modeling method is used to efficiently and accurately perform reliability analysis. A hydrokinetic turbine blade subjected to a vector of stochastic river flow loads is used to demonstrate the effectiveness of the proposed reliability analysis method.
机译：尾部相关性对于可靠性分析很重要，因为故障事件通常位于分布的尾部。多元ARMA模型和Karhunen-Loeve（KL）扩展方法用于多元随机负荷建模的常规做法无法捕获负荷之间的尾部相关性。本文提出了在相关非高斯随机负荷向量问题中，时变可靠性分析中负荷建模的藤蔓copula和自回归移动平均值（ARMA）的组合。单变量ARMA模型用于维持随时间变化的随机负载的相关性。 vine-copula模型不仅可以解释不同的单变量ARMA模型之间的相关性，而且可以解释不同ARMA模型的尾部相关性。所提出的Vine-ARMA模型能够在考虑尾部相关性的情况下灵活地对高维相关非高斯随机过程的矢量进行建模。为了克服计算工作中的挑战，最近开发的单循环克里格（SILK）替代建模方法用于有效和准确地执行可靠性分析。流体动力涡轮叶片受到随机河流量载荷矢量的作用，以证明所提出的可靠性分析方法的有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号