Abstract

To robustly estimate the fundamental mode component of prompt neutron decay constant α in a subcritical system, dynamic mode decomposition (DMD) is applied to time-series data obtained by the pulsed-neutron source (PNS) and Rossi-α methods. For the statistical uncertainty quantification of α by DMD, randomly sampled virtual data are used for the DMD procedure. The applicability of DMD is demonstrated by analyzing the experimental results by the PNS and Rossi-α methods, which are performed at the Kyoto University Critical Assembly (KUCA). When applying the DMD to the PNS and Rossi-α experimental data, a constant signal was added to the experimental data to remove the background constant component. The application results indicate that DMD enables one to robustly estimate the fundamental mode component of α in the PNS and Rossi-α methods.

摘要

为了稳健地估计亚临界系统中瞬发中子衰变常数 α 的基模分量，将动态模态分解 (DMD) 应用于通过脉冲中子源 (PNS) 和 Rossi-α 方法获得的时间序列数据。对于 DMD 对 α 的统计不确定性量化，随机采样的虚拟数据用于 DMD 程序。通过分析 PNS 和 Rossi-α 方法的实验结果证明了 DMD 的适用性，这些方法在京都大学关键大会 (KUCA) 上进行。当将 DMD 应用于 PNS 和 Rossi-α 实验数据时，将恒定信号添加到实验数据中以去除背景常数分量。应用结果表明，DMD 使人们能够在 PNS 和 Rossi-α 方法中稳健地估计 α 的基模分量。