A Dual-Path Physics-Informed Neural Network with Explicit Aging Dynamics for SOH Estimation of Lithium-ion Batteries

Clicks: 9
ID: 332084
2026
Free PDF
Article Quality & Performance Metrics
Overall Quality
Not rated
Combines reader engagement with the AI quality analysis. This article has not been analysed, so there is no overall score — reader engagement is measured and shown alongside.
AI Quality Assessment
Not analyzed
Readership in this journal
Emerging

Ranked #45 of 45 articles by views in journal of modern power systems and clean energy

Most read Least read

Bar heights use a square-root scale.

Mint this article as an NFT
Not yet minted

Create a permanent, verifiable on-chain record of this article on the Scimatic Network. The NFT is held in your Journament account, and you can withdraw it to your own wallet at any time.

5 SUSD one-off · no wallet required
Sign in to mint
Abstract
Abstract Accurate estimation of the state of health for lithium-ion batteries is crucial for battery management systems. To address the conflict between deep feature extraction and physical priors in single-path models, and the inability of fixed aging parameters to adapt to the dynamic degradation of lithium-ion batteries under real electric vehicle operating conditions, this paper proposes a dual-path physics-informed neural network for state-of-health prediction. First, the peak and area are extracted as aging features from the incremental capacity curves. Second, a dual-path parallel network architecture is constructed, comprising a feature attention-gated residual network prediction main branch and a dynamic parameter branch. Third, the universal differential equation paradigm is introduced to transform traditional fixed-parameter physical aging equations into a dynamic partial differential system driven by real-time health characteristics. Finally, the method is verified on the public Center for Advanced Life Cycle Engineering single-cell battery and self-assembled LBP battery pack dataset. Experimental outcomes demonstrate that the dual-path physics-informed neural network substantially enhances state-of-health estimation performance, with average mean absolute errors as low as 0.6920% and 0.2937%, respectively.
Reference Key
openalex_W7221364825 Use this key to autocite in the manuscript while using SciMatic Manuscript Manager or Thesis Manager
Authors Zhiqiang Lyu, Yunke Hu, Weiguo Kong, Geng Wang, Longxing Wu, Yuan Chen
Journal journal of modern power systems and clean energy
Year 2026
DOI
10.1093/ce/zkag067
URL
https://doi.org/10.1093/ce/zkag067
Keywords Keywords not found

Citations

No citations found. To add a citation, contact the admin at info@scimatic.org

Comments

No comments yet. Be the first to comment on this article.