TY  - EJOU
AU  - Lee, Kang-Woo 
AU  - Lee, Dong-Hee 
AU  - Kwon, Dae-Il 

TI  - A Hybrid Physics-Informed and Data-Driven Feature Framework with Explicit Correlation-Structure Embeddings for Early-Life Prognostics of Lithium-Ion Batteries
T2  - Computers, Materials \& Continua

PY  - 
VL  - 
IS  - 
SN  - 1546-2226

AB  - Early-life cycle-life prediction for lithium-ion batteries—estimating end-of-life from initial cycles—is valuable for rapid cell screening and battery health management. We investigate whether an explicit correlation-structure descriptor can complement physics-informed <i>ΔQ</i>-based indicators and generic early-cycle statistical features on the Severson 124-cell benchmark. We develop a lightweight hybrid framework that combines <i>ΔQ</i>-based health indicators, data-driven statistical features, and Laplacian Eigenmaps embeddings derived from a Pearson-correlation feature graph, with XGBoost used as the predictor. Across five feature configurations (<i>ΔQ Only, ΔQ + Statistics, Hybrid Append, VIF + Laplacian, and Integrated Laplacian</i>), we evaluate pointwise regression accuracy using RMSE and <i>R</i><sup>2</sup> together with PHM-style error-band measures RA@0.2, PH@0.1, and <i>α-λ</i>(0.15), computed on implied RUL trajectories induced by the early-life cycle-life estimate. On the Primary test domain, all four non-baseline configurations improved over <i>ΔQ Only</i>; Integrated Laplacian achieved the strongest RMSE/<i>R</i><sup>2</sup> pair (97.00 cycles, 0.8215), while Hybrid Append remained competitive (102.28 cycles, 0.8016) and improved RA@0.2 and PH@0.1 relative to <i>ΔQ + Statistics</i>. On the shifted Secondary domain, <i>ΔQ Only</i> gave the most favorable RMSE/<i>R</i><sup>2</sup> pair (267.82 cycles, 0.2275), whereas <i>Hybrid Append</i> and <i>VIF + Laplacian</i> improved selected error-band metrics. In an additional comparison against PCA, Random Projection, and Truncated SVD conducted at a matched 79-feature scale, with all transforms estimated from the training cells only, the graph-derived embedding remained competitive, but its margin over simpler reductions varied across splits. Taken together, these results support <i>Hybrid Append</i> as the main appended-structure configuration in this study, while indicating that the benefit of the correlation-structure descriptor is more visible in selected PHM-style error-band metrics than in uniformly improved pointwise accuracy.
KW  - Lithium-ion battery; early-life cycle life prediction; remaining useful life (RUL); prognostics and health management (PHM); <i>ΔQ</i>-based health indicators; hybrid feature engineering; graph Laplacian embedding; prognostic metrics

DO  - 10.32604/cmc.2026.081667