Driving Design with Cognitive Simulation

Jayson Peterson | 21-ERD-028

Executive Summary

We are developing new solutions to produce optimal designs using complex multiphysics codes to address problems in inertial confinement fusion and high-energy-density science. By combining novel design strategies with a new class of cognitive simulation models, the research will enable optimal experiment designs for stockpile stewardship and other national security missions.

Publications, Presentations, and Patents

Gillette, A. February 2021. “Delaunay-Guided Neural Network Design.” Colorado State Applied Math Seminar. LLNL-PRES-819811.

Gillette, A. March 2021. “Delaunay-Based Assessment of Variational Autoencoders.” SIAM Conference on Computational Science and Engineering. LLNL-PRES-819941.