With the large-scale application of electric vehicles and distributed energy storage, lithium-ion batteries are playing an increasingly prominent role in power and energy networks. Recent studies reveal critical gaps in battery fast-charging research: (i) the absence of unified frameworks that co-design grid and infrastructure constraints with degradation and thermal limits, (ii) limited use of real-time or early-life degradation predictors in adaptive charging, and (iii) insufficient balancing of charging speed, economics, and long-term battery health. To address these challenges, this paper proposes a health-aware adaptive multi-stage constant current–constant voltage (MS-CC-CV) fast-charging framework that integrates thermo-aging limits and elastic-net regression predictors capable of forecasting cycle life with 87% accuracy from the first 100 cycles. A Grey Wolf Optimization (GWO) algorithm designs charging profiles while explicitly embedding thermal and safety constraints into decision-making. Experimental validation on APR18650M1B and LG-EBM26R cells demonstrates cycle-life improvements of 28% and 17.23%, respectively, compared with 1 C CC–CV charging, with competitive or reduced charging times. An additional validation protocol achieved 524 cycles within ~50 minutes, confirming the practical potential of the framework. By coupling predictive health indicators with adaptive charging optimization, this work provides a scalable and grid-ready solution for degradation-conscious charging in electric mobility and stationary storage systems
Bose et al. (Sun,) studied this question.