Development of Genetic Algorithm Based Multi-Objective Plant Reload Optimization Platform

Kim, Junyung; Abdo, Mohammad; Wang, Congjian; Cho, Yong-Joon

Permalink

https://hdl.handle.net/2142/121858 Copy

Description

Title

Development of Genetic Algorithm Based Multi-Objective Plant Reload Optimization Platform

Author(s)

• Kim, Junyung
• Abdo, Mohammad
• Wang, Congjian
• Cho, Yong-Joon

Issue Date

2023

Keyword(s)

• Genetic algorithm
• Plant reload optimization
• Multi-objective optimization
• Non-dominated sorting genetic algorithm II

Abstract

The United States (U.S.) Department of Energy (DOE) Light Water Reactor Sustainability (LWRS) Program Risk-Informed Systems Analysis (RISA) Pathway Plant Reload Optimization Project aims to develop an integrated, comprehensive platform offering an all-in-one solution for reload evaluations with a special focus on core design optimization. Fuel reload optimization is a multi-physics problem that needs to consider core design, fuel performance, and system safety. The NSGA-II (non-dominated sorting genetic algorithm II), a variant of genetic algorithm (GA) was selected as the foundation for the optimization platform for the multi-objective optimization problems (MOOPs) to optimize multiple objectives, such as fuel cycle length, enrichment, and burnable poisons with multiple constraints, including core design limits and system safety parameters. Idaho National Laboratory’s Risk Analysis and Virtual Environment (RAVEN) is being used as the workflow manager and a fuel reload optimization platform. RAVEN controls the perturbations of input decks to all the physics codes in neutronics, fuel performance, and safety analyses via generic and specialized built-in code interfaces, parses inputs and outputs, and performs post-processing of the simulation results. The recent activity progressed for the GA optimization platform by implementing more evolutionary operations mechanisms (i.e., fitness functions, parent selection, crossover, mutation, and survivor selection). This paper summarizes genetic-algorithm- based multi-objective fuel reload optimization activities, specifically: developing the non-dominated sorting genetic algorithm II optimizer in the RAVEN and demonstrating and validating the developed NSGA II optimizer using benchmark optimization problems.

Type of Resource

text

Language

eng

Handle URL

https://hdl.handle.net/2142/121858

Owning Collections