Speaker
Description
\documentclass{style/nseJournal}
\usepackage{enumitem}
\usepackage{xcolor}
\usepackage{hyperref}
\usepackage{cleveref}
\usepackage{graphicx}
\usepackage{rotating}
\usepackage{ulem}
\newcommand{\beff}{$\beta_\mathrm{eff} $}
\newcommand{\keff}{$k_\mathrm{eff}$}
\begin{document}
\title{EUCLID: A New Approach to Improve Nuclear Data Coupling Optimized Experiments with Validation using Machine Learning}
\addAuthor{\correspondingAuthor{Jesson Hutchinson}}{a}
\correspondingEmail{jesson@lanl.gov}
\addAuthor{Jennifer Alwin}{a}
\addAuthor{Alex{and}er Clark}{a}
\addAuthor{Theresa Cutler}{a}
\addAuthor{Michael Grosskopf}{a}
\addAuthor{Wim Haeck}{a}
\addAuthor{Michal Herman}{a}
\addAuthor{Noah Kleedtke}{a}
\addAuthor{Juliann Lamproe}{a}
\addAuthor{Robert Little}{a}
\addAuthor{Isaac Michaud}{a}
\addAuthor{Denise Neudecker}{a}
\addAuthor{Michael Rising}{a}
\addAuthor{Travis Smith}{a}
\addAuthor{Nicholas Thompson}{a}
\addAuthor{Scott Vander Wiel}{a}
\addAffiliation{a}{Los Alamos National Laboratory\ P.O. Box 1663, MS B228, Los Alamos, New Mexico 87545}
%\addKeyword{None}
\titlePage
\section{Abstract}
\label{sec:Intro}
Unconstrained physics spaces arise between nuclear data due to imprecise knowledge or missing differential experimental data and theory combined with the fact that validation with integral experiments cannot uniquely identify one nuclear-data value as driving the difference between simulated and experimental values of these experiments. Compensating errors of nuclear data can easily hide within these unconstrained physics spaces. This can potentially have large impacts on application calculations that do not closely resemble validation experiments we validate our nuclear data against.
The EUCLID project (Experiments Underpinned by Computational Learning for Improvements in Nuclear Data) aims to resolve part of these compensating errors by performing an integral experiment optimized to better constrain nuclear data. In addition to that, the project will adjust ENDF/B-VIII.0 nuclear data with respect to this new experiment to better constrain nuclear data.
EUCLID includes four major components: (1) simulations, (2) experiments, (3) machine learning, and (4) nuclear data validation/adjustment. One major component of EUCLID is the evaluation of several integral measurement methods (criticality, reaction rate ratios, neutron noise, etc.). Regarding simulations, new capabilities are being developed to estimate sensitivity coefficients for all of these methods. Regarding experiments, the National Criticality Experiments Research Center (NCERC) will be utilized to perform measurements. These experiments will be optimized using Bayesian optimization. The measured responses (along with the simulated responses and sensitivities) will be utilized via Random Forest to perform nuclear data adjustment. This will include requirements to obey physics constraints and will result in adjusted nuclear data.
\begin{figure}[htb!]
%\vspace{-20pt}
\centering\includegraphics[width=6in]{figures/EUCLIDimage.pdf}
%\vspace{-5pt}
\caption{Unconstrained physics spaces within nuclear data can adversely impact application simulations, but cannot be resolved using current techniques. EUCLID aims to provide a capability to help understand such unconstrained spaces. In order to achieve this, we utilize machine learning with simulated sensitivities in MCNP of experiments and measured results to augment expert identification of areas where compensating errors may be resolved. The design of new experiments will then be optimized using machine learning. After the new experiments are completed, the new data will be used to produce adjusted libraries.}
\label{Fig_EUCLID}
%\vspace{-25pt}
\end{figure}
\newpage
\section*{Acknowledgments}
\label{sec:acknowledgements}
\textcolor{black}{Research reported in this publication was supported by the U.S. Department of Energy LDRD program at Los Alamos National Laboratory. The National Criticality Experiments Research Center (NCERC) is supported by the DOE Nuclear Criticality Safety Program, funded and managed by the National Nuclear Security Administration for the Department of Energy. This work was supported by the US Department of Energy through the Los Alamos National Laboratory. Los Alamos National Laboratory is operated by Triad National Security, LLC, for the National Nuclear Security Administration of the US Department of Energy under Contract No. 89233218CNA000001.}
\end{document}