MetaHeuristic Feature Selection for Energy Group Optimization and Analysis
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
Energy discretization is a crucial component of deterministic neutron transport simulations. Metaheuristic (MH) optimizers are effective algorithms to determine group structures that maximize both solution accuracy and computational efficiency. This project establishes a framework for optimizing group structures for PARTISN simulations using the Python library MEALPY. Group structure optimization is formulated as a binary feature selection problem, and results are investigated with permutation and material importance techniques to determine physically relevant energy bounds. We conclude that MH optimizers find group structures that drastically improve flux calculations while preserving k-effective accuracy. Further, we find that individual energy bounds are not necessarily physically relevant, but rather specific energy ranges are.
- Research Organization:
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
- Sponsoring Organization:
- USDOE Office of Nuclear Energy (NE); USDOE National Nuclear Security Administration (NNSA)
- DOE Contract Number:
- 89233218CNA000001
- OSTI ID:
- 2588817
- Report Number(s):
- LA-UR--25-29285
- Country of Publication:
- United States
- Language:
- English
Similar Records
Particle Swarm Optimisation for group structure optimization for radiotherapy shielding
Metaheuristic Optimization Tool
Applications of the renormalization group
Conference
·
Fri Jul 01 00:00:00 EDT 2022
·
OSTI ID:23203862
Metaheuristic Optimization Tool
Technical Report
·
Sat Feb 29 23:00:00 EST 2020
·
OSTI ID:1608209
Applications of the renormalization group
Journal Article
·
Mon Sep 25 00:00:00 EDT 2006
· AIP Conference Proceedings
·
OSTI ID:20888292