Enhancing Battery Health Monitoring in Battery Assembly Digital Twin (BATCAT)

Wissenschaftlicher Mitarbeiter
Artificial Intelligence and Cybersecurity (AICS) |  Production Systems (PROSYS)
Institute for Smart System Technologies (SST) | Transportation Informatics Group  (TIG )

Abstract:

This work is part of the BatCAT project, which aims to advance battery manufacturing through the development of digital twins by integrating multi-physics models, machine learning, and explainable AI. In Work Package 5, we focus on creating machine learning models to monitor battery health, specifically targeting key metrics such as State of Health (SOH) and Remaining Useful Life (RUL). Current efforts include creating and testing hybrid models like Cellular Neural Networks (CeNNs), Transformers, and Autoencoders, which are designed to capture both spatial and temporal patterns in battery systems. To ensure transparency and trust in the predictions, we integrate explainable AI methods, while maintaining high performance. This work contributes to reliable and interpretable tools for monitoring battery health in the battery assembly digital twin (BatCat).