Lithium-ion cells with nearly identical capacity state of health (SOH) can still differ substantially in load response. We introduce DN-State Channel, a categorical label that reports which measured health channel — capacity or an apparent load-response resistance — currently limits a cell, together with DN-State, a continuous joint-condition descriptor. The framework is training-free and was applied without retuning to three public datasets. On NASA cycling data (33 standard-cycling cells and 4 random-walk cells), DN-State tracked SOH (mean |Spearman| 0.875 and 0.760–0.958), and the dominant channel reversed between loading regimes: capacity-limited under standard cycling (82.2% of cycles) and resistance-limited under random-walk loading (72–95% per cell). On the UCF accelerated-life dataset (22 cells), channel-switching activity separated cells under identical nominal loading into two clusters, rose near documented load transitions above about 3 A in 6 of 8 cells, and tie frequency tracked prior-life usage in second-life cells. On the PulseBat retired-cell dataset (203 cells, three chemistries), labels fixed at 40% SOC were tested against a withheld 1.5C pulse within SOH-matched pairs. The higher-resistance-channel cell showed the poorer held-out response in 93.8–100% of pairs (paired p between 3.05×10⁻⁵ and 3.23×10⁻⁴). The ordering held across 5–50% SOC and up to 2.5C in two cohorts and from 10% SOC in the third. We report negative and cohort-limited results with equal visibility and propose a multi-dimensional extension of the limiting-channel label for end-of-life triage.
No takes yet. Share an insight, caveat, or question.
Duc Vinh Nguyen (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: