Nuclear Engineering ETDs

Publication Date

Fall 12-13-2024

Abstract

There is an increasing interest in developing and deploying small modular nuclear reactors (SMRs) and micro modular reactors (MMRs). They could be operated with minimal on-site personnel or remotely with a high degree of autonomy. This can be enabled through the use and application of Artificial Intelligence (AL) and Machine Learning (ML) algorithms and methods. The objective of the present work is to train machine learning algorithms for remote control mimicking the transient startup of the Very-Small, Long-LIfe, Modular (VSLLIM) microreactor. This walk-away safe microreactor design and a fully integrated transient model have been developed at the University of New Mexico’s Institute for Space and Nuclear Power Studies (UNM-ISNPS). This work investigates using the Supervised Learning (SL) and Reinforcement Learning (RL) paradigms to train Artificial Neural Networks (ANNs) to manage the movement of the control rods in the VSLLIM microreactor during simulated startup transients. The trained ANNs attempt to predict the displacement of the control rods in the reactor core for a smooth startup to nominal, full-power steady state condition.

Keywords

Machine Learning, Reinforcement Learning, Microreactor, Reactor Control, Transient Startup, Soft Actor-Critic

Document Type

Thesis

Language

English

Degree Name

Nuclear Engineering

Level of Degree

Masters

Department Name

Nuclear Engineering

First Committee Member (Chair)

Mohamed S. El-Genk

Second Committee Member

Christos Christodoulou

Third Committee Member

Timothy M. Schriener

Download

Share

COinS