Utilizing Deep Learning for Crystal System Classification in Lithium - Ion Batteries
Yibo Yin, Guokun Xu, Ying Xie, Yang Luo, Zibu Wei, Zhengning Li
Abstract
Lithium-ion (Li-ion) batteries are pivotal in energy storage, powering diverse applications from portable electronics to electric vehicles. Optimizing Li-ion battery performance relies on understanding the crystal system properties of constituent materials, notably cathodes. This paper proposes a novel approach using Deep Neural Networks (DNNs) for multi-class classification of Li-ion silicate cathode crystal systems. Previous research underscores crystal chemistry's importance and the potential of machine learning in Li-ion battery materials. However, existing methodologies face challenges in accurately capturing material complexities. Our DNN-based model aims to address these limitations, offering improved predictive performance for crystal system classification.