SeqBattNet: A Discrete-State Physics-Informed Neural Network with Aging Adaptation for Battery Modeling

Abstract

SeqBattNet is a discrete-state physics-informed neural network for battery modeling with built-in aging adaptation. An HRM-GRU encoder derives cycle-specific aging parameters, while an equivalent-circuit decoder combines those parameters with the input current to predict terminal voltage during discharge. The model uses only three basic battery parameters and demonstrates robust, efficient voltage prediction across multiple public battery datasets.

@article{tran2025seqbattnet,
  title={SeqBattNet: A Discrete-State Physics-Informed Neural Network with Aging Adaptation for Battery Modeling},
  author={Tran, Khoa and Trinh, Hung-Cuong and Nguyen, Vy-Rin and Nguyen-Thoi, T. and Nguyen-Thai, Vin},
  journal={arXiv preprint arXiv:2509.17621},
  year={2025},
  doi={10.48550/arXiv.2509.17621}
}