Education, Science, Technology, Innovation and Life
Open Access
Sign In
  1. Home
  2. Journals
  3. Advances in Computer, Signals and Systems
  4. Vol 10, Issue 2, 2026

PL-LSTM-Based State of Health Estimation and TTE Prediction for Lithium-Ion Batteries

Download as PDF

DOI: 10.23977/acss.2026.100204 | Downloads: 1 | Views: 124

Author(s)

Yeqi Zhang 1, Zekai Yu 1, Wenyue Wang 1, Yanyan Wu 2,3

Affiliation(s)

1 Big Data College, Fuzhou College of Foreign Studies and Trade, Fuzhou, China
2 College of Artificial Intelligence, Ningbo University of Finance and Economics, Ningbo, China
3 Faculty of Data Science, City University of Macau, Macau, China; Key Laboratory of Data Science and Intelligent Computing, Fuzhou University of International Studies and Trade, Fuzhou, Fujian, China

Corresponding Author

Yanyan Wu

ABSTRACT

This study proposes a hybrid model, PL-LSTM, that integrates a Physics-Informed Neural Network (PINN) with a Long Short-Term Memory (LSTM) network for estimating the State of Health (SOH) and predicting the Time to End-of-Life (TTE) of lithium-ion batteries. The model extracts dynamic temporal features within an extremely short time window and embeds physical constraints—derived from the first-order RC equivalent circuit and SOC evolution equation—into a composite loss function, thereby achieving an organic integration of data-driven learning and mechanistic modeling. Experimental results demonstrate that PL-LSTM significantly outperforms traditional LSTM models on the CALCE battery aging dataset and the GreenHub dynamic load dataset. For SOH estimation, PL-LSTM effectively reduces MAE and RMSE, accurately capturing nonlinear degradation and abrupt capacity changes; for TTE prediction, the model reveals the dual-driving mechanism of capacity decay and internal resistance growth, validating the amplifying effect of aging on lifespan reduction under high-load scenarios. Overall, PL-LSTM enhances the accuracy and robustness of battery state estimation while providing interpretable physical insights, offering a new technical pathway for intelligent and reliable battery management in electric vehicles, energy storage systems, and mobile devices.

KEYWORDS

Lithium-ion battery, SOH estimation, TTE prediction, PINN, LSTM, physical constraints, battery management system

CITE THIS PAPER

Yeqi Zhang, Zekai Yu, Wenyue Wang, Yanyan Wu. PL-LSTM-Based State of Health Estimation and TTE Prediction for Lithium-Ion Batteries. Advances in Computer, Signals and Systems (2026). Vol. 10, No. 2, 30-41. DOI: http://dx.doi.org/10.23977/acss.2026.100204.

REFERENCES

[1] Yao Fang, Yang Xiaona, Li Qian, et al. Cooperative Estimation of SOC and SOH in Lithium-Ion Batteries Considering Coupling and Correlation Characteristics [J]. Power Electronics Technology, 2024, 58(11): 85–90.
[2] Pereira, R., Matalonga H ,Couto M , et al. GreenHub: a large-scale collaborative dataset to battery consumption analysis of android devices[J].Empirical Software Engineering,2021,26(3):1-55.
[3] DOYLE M, FULLER T F, NEWMAN J. Modeling of Galvanostatic Charge and Discharge of the Lithium/Polymer/Insertion Cell[J]. Journal of The Electrochemical Society, 1993, 140(6): 1526-1533.
[4] S. Liu et al., "State of Charge Estimation of Lithium Battery Based on Improved Unscented Kalman filter," 2024 IEEE 10th International Conference on Underwater System Technology: Theory and Applications (USYS), Xi'an, China, 2024, pp. 1-6.
[5] A. J B ,René B ,David S , et al. State of charge and state of health diagnosis of batteries with voltage-controlled models[J].Journal of Power Sources,2022,544.
[6] Khan M S, Maddipatla S, Pecht M. A review of float charging in lithium-ion batteries: Degradation mechanisms, influencing factors, and optimization strategies[J]. Journal of Power Sources, 2026, 616: 235861.
[7] Mei Yuankun. A Study on Vehicle Intention Recognition on Expressways Using PL-LSTM [J]. Modern Information Technology, 2026, 10(01): 81-85.
[8] Guo Shengpin, Wang Huiming. Deep Learning Research on Concrete Failure Criteria Based on Physical Information Neural Networks [J]. Concrete, 2024, (09): 28–34.
[9] An Peng, Cao Danping, Zhao Baoyin, et al. A Study on Reservoir Property Parameter Prediction Methods Based on LSTM Recurrent Neural Networks [J]. Advances in Geophysics, 2019, 34(5): 1849–1858.
[10] WANG Q, GAO T, LI X C. SOC estimation of Lithium-ion battery based on equivalent circuit model with variable parameters[J]. Energies, 2022, 15(16):5829.

Subscription
E-Mail Alert
Most popular papers
Call for Special Issue
Publication Ethics & OA Statement
Plagiarism Policy
Contact Us
Downloads: 49380
Visits: 1128594

Sponsors, Associates, and Links

All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.