Abstract
Large-scale lithium-ion battery (LiB) applications significantly promote marine transport decarbonization. However, inevitable marine transport swaying conditions markedly deteriorate the health of batteries. There are no quantitative methods to elucidate unique degradation behaviors and their unknown interdependencies. In this context, this article proposes a physics-informed battery aging model (PIBM). First, the dominant aging patterns of batteries affected by the external mechanical stress are revealed through a laboratory-scale experimental system with standard marine transport swaying conditions. Then, a universal pseudo-2-D (P2D) electrochemical model combining coupled degradation mechanisms is introduced to quantify the particle-level state degradation. Subsequently, a general data-driven strategy that respects physical causality is embedded into the modified P2D model. This strategy can approximate the finite-order partial differential equations (PDEs) and efficiently resolve the full-life Li+ diffusion dynamics. Compared with the conventional hybrid data-driven model, the proposed method quantifies the health degradation of batteries from the particle to the cell level and demonstrates more accurate and faster performance.
Funding Information
  • Smart Grid-National Science and Technology Major Project (2025ZD0805401)
  • National Natural Science Foundation of China (5257070416)
  • National Natural Science Foundation of China (202506050093)

This publication has 48 references indexed in Scilit: