Nuclear engineering doctoral student John Pevey was selected as a recipient of the 2022 Mark Mills Award for his paper titled “Neural Network Acceleration of Genetic Algorithms for the Optimization of A Coupled Fast/Thermal Nuclear Experiment.”
Pevey co-wrote the piece with his co-advisors, Assistant Professor Vladimir Sobes and Postelle Professor, Chancellor’s Professor, and Department Head Wes Hines. Pevey wrote this paper with the purpose of accelerating the optimization of the design of the Fast Neutron Source (FNS).
As he received notice of the award he was extremely grateful for his advisors and committee members that supported him throughout the research process.
“I would like to specifically acknowledge Dr. Hines, Dr. Sobes, Dr. Chvala, Dr. Ilas, and Dr. Bogetic, as well as the entire graduate and undergraduate FNS team here at UT,” he said.
In the paper, he compared two optimizations of the FNS. The first was the standard genetic algorithm and the second was a standard genetic algorithm where the data produced during the calculation was used to produce complex non-linear functions called neural networks.
The neural network in the second optimization was able to predict the nuclear characteristics of new designs. This allowed for more potential FNS designs to be produced.
The award, presented by the American Nuclear Society, consists of an engraved plaque and a monetary prize of $500. The recipient must be one who has authored an original paper as defined by the ANS constitution and bylaws concerning papers, published in or submitted to a meeting proceeding, that is judged by the ANS Awards Committee to contribute substantially to the advancement of science and engineering related to atomic nuclei. Typically, the award is presented annually at the ANS winter meeting.
Pevey will graduate this December. He is now working as a neutronic analyst at X-energy located in Washington, DC.