退役锂离子电池数量逐年增加。对退役电池进行梯次利用可有效提高资源利用效率。然而，退役电池组中电池单体之间的性能差异较大，这种不一致性会加剧梯次利用过程中的性能衰减、安全风险及热失控隐患。现有评估方法多基于单一参数或简单统计指标，难以全面反映多维参数间的复杂关联，且缺乏针对退役电池组的系统性量化体系。针对上述问题，本文提出一种基于多维参数融合的退役电池组不一致性综合量化方法。首先，选取容量衰减率、欧姆内阻、极化内阻及剩余使用寿命作为核心表征参数，构建电池组的多维参数集。其次，从离散度、不确定性和鲁棒性三个层面，分别采用基尼均值差、泰勒指数及无量纲四分位距作为统计指标，并通过加权融合构建综合不一致性指数（CI）。基于统计理论推导CI的可接受阈值，为梯次利用决策提供量化依据。进一步，针对不合格电池组，提出按权重最高参数进行排序重组并剔除劣质单体的优化策略。利用多维威布尔分布根据退役锂电池普遍的7.5% ～ 20%参数偏差，生成100个电池组（每组1000个单体）的模拟数据开展验证。结果表明：CI值能够有效区分合格与不合格电池组；按侧重参数重组可使CI值降低22.6%～28.9%，显著提升一致性。该方法为退役电池的精准拆解、优化重组及高效梯次利用提供了系统的理论支撑与实用工具。

The number of retired lithium-ion batteries is increasing year by year. Cascade utilization of retired batteries can effectively improve resource utilization efficiency. However， the performance differences among individual cells within retired battery packs are substantial， and such inconsistency can exacerbate performance degradation， safety risks， and thermal runaway hazards during cascade utilization. Existing evaluation methods are mostly based on single parameters or simple statistical indicators， which are difficult to comprehensively reflect the complex correlations among multi-dimensional parameters， and lack a systematic quantification framework specifically for retired battery packs. To address these issues， this paper proposes a comprehensive quantification method for inconsistency in retired battery packs based on multi-dimensional parameter fusion. First， capacity fade rate， ohmic internal resistance， polarization internal resistance， and remaining useful life are selected as core characterization parameters to construct a multi-dimensional parameter set for the battery pack. Second， from the three perspectives of dispersion， uncertainty， and robustness， the Gini mean difference， Theil index， and dimensionless interquartile range are adopted as statistical indicators， respectively， and a comprehensive inconsistency index (CI) is constructed through weighted fusion. The acceptable threshold of CI is derived based on statistical theory， providing a quantitative basis for cascade utilization decisions. Furthermore， for unqualified battery packs， an optimization strategy is proposed that ranks and regroups cells according to the parameter with the highest weight while eliminating inferior cells. A multi-dimensional Weibull distribution is employed to generate simulated data for 100 battery packs (1，000 cells per pack) based on the typical parameter deviation of 7.5% to 20% in retired lithium batteries for validation. The results demonstrate that the CI value can effectively distinguish qualified from unqualified battery packs， and regrouping according to the prioritized parameter can reduce the CI value by 22.6% to 28.9%， significantly improving consistency. The proposed method provides systematic theoretical support and a practical tool for precise disassembly， optimized regrouping， and efficient cascade utilization of retired batteries.