Volume 10,Issue 8
LSTM-Based Operating-State Forecasting and Risk Warning for Small-Scale Battery Energy Storage Systems Using Multi-Source Aging Data
Senseless computing resources are needed to achieve reliable forecasting in small-scale battery energy storage. In this research, the authors present an LSTM-based framework to normalize NASA and CALCE aging data, separate the data for battery-level prediction, predict 5-cycle-ahead state of health (SOH), and transform residual and health signals to warnings. In this work, LSTM was tested against the NASA and CALCE datasets using a common seven-model protocol and obtained competitive cross-cell performance (NASA: MAE 0.01219, RMSE 0.01619; CALCE: MAE 0.01149, RMSE 0.02331). Controlled anomaly tests detected all severe abrupt drops, and low-SOH risk was separated from model unexplained deviations by replay on CALCE CS2_38. Results of a rerun with an independent CPU were consistent with the original rankings of models and the statistical interpretation. The result is that the framework offers a simple and repeatable foundation for battery-state monitoring, but validation of the field-fault is still required.
[1] Yang K, Zhang L, Zhang Z, et al., 2023, Battery State of Health Estimate Strategies: From Data Analysis to End-Cloud Collaborative Framework. Batteries, 9(7): 351.
[2] Hochreiter S, Schmidhuber J, 1997, Long Short-Term Memory. Neural Computation, 9(8): 1735–1780.
[3] Zhao S, Luo L, Jiang S, et al., 2023, Lithium-Ion Battery State-of-Health Estimation Method Using Isobaric Energy Analysis and PSO-LSTM. Journal of Electrical and Computer Engineering, 2023: 5566965.
[4] Feng S, Song M, Lin Y, et al., 2024, Convolutional Neural Network-Long Short-Term Memory-Based State of Health Estimation for Li-Ion Batteries Under Multiple Working Conditions. Energy Technology, 12(2): 2301039.
[5] Lin M, Hu D, Meng J, et al., 2025, Transfer Learning-Based Lithium-Ion Battery State of Health Estimation with Electrochemical Impedance Spectroscopy. IEEE Transactions on Transportation Electrification, 11(3): 7910–7920.
[6] Paszke A, Gross S, Massa F, et al., 2019, PyTorch: An Imperative Style, High-Performance Deep Learning Library. Advances in Neural Information Processing Systems, 32.
[7] NASA 2026, Li-Ion Battery Aging Datasets, visited on 29 Jul 2026, https://data.nasa.gov/dataset/li-ion-battery-aging-datasets
[8] Center for Advanced Life Cycle Engineering, n.d., CALCE Battery Data: CS2 Battery, University of Maryland, visited on 29 Jul 2026, https://calce.umd.edu/battery-data.
[9] He W, Williard N, Osterman M, et al., 2011, Prognostics of Lithium-Ion Batteries Based on Dempster-Shafer Theory and the Bayesian Monte Carlo Method. Journal of Power Sources, 196(23): 10314–10321.
[10] Xing Y, Ma E, Tsui K, et al., 2013, An Ensemble Model for Predicting the Remaining Useful Performance of Lithium-Ion Batteries. Microelectronics Reliability, 53(6): 811–820.
[11] Williard N, He W, Osterman M, et al., 2013, Comparative Analysis of Features for Determining State of Health in Lithium-Ion Batteries. International Journal of Prognostics and Health Management, 4(1): 1–7.
