首页> 外文会议>International conference on the physics of reactors;PHYSOR 08 >A generalized KPSS test for stationarity detection in Monte Carlo eigenvalue problems
【24h】

A generalized KPSS test for stationarity detection in Monte Carlo eigenvalue problems

机译:蒙特卡洛特征值问题中平稳性检测的广义KPSS检验

获取原文

摘要

Progress with source convergence diagnostics over the past several years has led to the development of new techniques for identifying a stationary source for Monte Carlo eigenvalue problems. We have adapted a new statistics based diagnostic comparing the residuals of a linear regression of a data series to its expected distribution. This test is called the Generalized KPSS test for stationarity. We have implemented this test and others into a modified version of KENO V.a. Using the OECD/NEA criticality safety benchmark problem 4 -Array of Interacting Spheres, the KPSS test is compared with other stationarity detection tests such as the information theoretic approach and a limiting distribution approach. Results indicate favorable diagnostic capability in idealized cases. Three different cases with 50 replications are examined. Depending on the problem case and combination of different diagnostics, we have identified 98% to 100% of the non-stationary cases. We conclude that the best action for Monte Carlo eigenvalue codes is to combine different diagnostic tests.
机译:在过去的几年中,源收敛诊断技术取得了进步,从而导致了用于确定蒙特卡洛特征值问题的固定源的新技术的发展。我们采用了一种新的基于统计的诊断方法,将数据序列的线性回归残差与其预期分布进行比较。该测试称为平稳性的通用KPSS测试。我们已经将此测试和其他测试实施到KENO V.a的修改版本中。使用OECD / NEA临界安全基准问题4-相互作用球阵列,将KPSS测试与其他平稳性检测测试(例如信息理论方法和极限分布方法)进行比较。结果表明在理想情况下具有良好的诊断能力。研究了三种不同的情况,其中有50个重复。根据问题案例和不同诊断方法的组合,我们确定了98%到100%的非平稳案例。我们得出结论,蒙特卡洛特征值代码的最佳操作是组合不同的诊断测试。

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