Updates to the Variational Auto-encoder for crack parametrization

作者：Sandia National Laboratories (SNL-NM), Albuquerque, NM (United States), Norman Jost, Solar Energy Technologies Program, Ojas Sanghi, Brandon Byford, Benjamin Pierce, Emma Cooper, Isaiah Deane, Jennifer Braid · 年份：2025 · DOI：10.2172/3020983 · 研究领域：Applied mathematics、Mathematics、Mathematical analysis、Physics、Computer science

New materials and designs for photovoltaic modules providing low-cost energy Multi-Scale, Multi-Physics Model Sandia PV This work will develop modular, open-source model and analysis components including crack detection workflow and parameterization for quantitative inspection of large EL large datasets.These tools will allow users to quickly and accurately assess the extent and types of cracking in their modules.Measured statistical distributions of crack parameters, together with the imposed stress and electrical properties will be used to generate models to predict future crack behavior and power loss.