Posts by Collection

portfolio

Research Portfolio

배터리 PHM 및 Edge AI 연구 포트폴리오

publications

Enhancing Battery SOH Prediction with Physics-Informed Neural Networks in Data-Scarce Environments

Published in Energy, 2025

Demonstrated superior performance compared to existing models in data-scarce environments by integrating physics-based learning.

Recommended citation: Y. Seo, T. Kim, S. Barde. (2025). "Enhancing Battery SOH Prediction with Physics-Informed Neural Networks in Data-Scarce Environments." Energy.
Download Paper

Hybrid Compression for Accurate End of Life Prediction on Edge Battery Management System

Published in IEEE Global Reliability & PHM Conference, 2025

Developed lightweight model enabling real-time battery EOL estimation on edge devices.

Recommended citation: T. Kim, Y. Seo, S. Barde. (2025). "Hybrid Compression for Accurate End of Life Prediction on Edge Battery Management System." IEEE Global Reliability & PHM Conference.
Download Paper

Robust SOH Prediction for Lithium-Ion Batteries via ProbSparse Informer Architecture

Published in IEEE Global Reliability & PHM Conference, 2025

Best Paper Award

Recommended citation: Y. Seo, T. Kim, S. Barde. (2025). "Robust SOH Prediction for Lithium-Ion Batteries via ProbSparse Informer Architecture." IEEE Global Reliability & PHM Conference.
Download Paper

Edge-compatible SOH Estimation for Li-ion Batteries via Hybrid Knowledge Distillation and Model Compression

Published in Journal of Energy Storage, 2025

Achieved 83.5× model compression (451KB → 5.4KB) while maintaining accuracy. Implemented lightweight model capable of real-time inference on Raspberry Pi 4B. Improved throughput by 18.7× and inference speed by 13.1× compared to baseline.

Recommended citation: T. Kim, Y. Seo, S. Barde. (2025). "Edge-compatible SOH Estimation for Li-ion Batteries via Hybrid Knowledge Distillation and Model Compression." Journal of Energy Storage.
Download Paper