Dynamic modeling of lithium-ion battery degradation using data-driven and physics-informed method

Authors

  • Daniel Santoso Department of Electronic and Computer Engineering, Universitas Kristen Satya Wacana
  • Muhamad Dzaky Ashidqi Department of Electrical Engineering and Information Technology, Universitas Gadjah Mada Department of Electrical Engineering, Universitas Sains Indonesia http://orcid.org/0009-0003-4584-0441

DOI:

https://doi.org/10.22441/sinergi.2026.1.013

Keywords:

Capacity degradation, Equivalent Circuit Model, Lithium-Ion Battery, Neural Network,

Abstract

Accurate realtime prediction of lithiumion battery (LIB) capacity degradation is essential for embedded batterymanagement systems. Equivalent circuit models (ECMs) run quickly but lose accuracy over time, whereas purely data-driven networks achieve high precision at a high computational cost. This study introduces a physicsinformed neural network (PINN) that embeds the differential equations of a firstorder Thevenin ECM directly into the loss function. Using only terminal voltage and current as inputs, the network simultaneously estimates internal resistance, polarization resistance, polarization capacitance, opencircuit voltage, and capacity loss. The model was trained and evaluated over 300 chargedischarge cycles of a 18650 lithium-ferrous phosphate (LFP) cell. The resulting capacity degradation estimation achieved a root mean squared error (RMSE) of 0.012and a mean absolute percentage error (MAPE) of 0.974%, surpassing a neural ordinary differential equation baseline with RMSEof0.215. The trained network contains 261 parameters, requires 0.6 ms per sample for inference, and consumes 49 MB of memory. This computation cost is far lower than that of a long shortterm memory (LSTM) benchmark with comparable accuracy. In addition, the proposed model maintains its accuracy under limited dataset conditions. With a fourfold smaller training set, the PINN maintained an RMSE of 0.023, whereas the LSTM error increased to 0.72. The results demonstrate that lightweight neural networks guided by physics-based constraints can provide reliable, real-time health estimation on resourcelimited hardware.

Downloads

Download data is not yet available.

Downloads

Additional Files

Published

2026-01-07

How to Cite

[1]
D. Santoso and M. D. Ashidqi, “Dynamic modeling of lithium-ion battery degradation using data-driven and physics-informed method”, Sinergi, vol. 30, no. 1, pp. 135–146, Jan. 2026.

Issue

Section

Articles

Similar Articles

<< < 

You may also start an advanced similarity search for this article.