Remaining Useful Life Prediction of Lithium-Ion Batteries Using Monotone Decomposition

Accurately predicting the remaining useful life (RUL) of lithium-ion batteries is vital for efficient equipment health management. Throughout the aging process, the battery capacity exhibits nonlinear behavior, with intermittent capacity regeneration phenomena causing sudden increments between consecutive cycles, posing challenges for modeling and prediction. Despite the frequent use of empirical mode decomposition (EMD) to decompose capacity series, most EMD-based RUL prediction methods encounter limitations including end effects, information leakage issues, and a lack of uncertainty quantification. To address these challenges, we introduce a novel RUL prediction framework, MonoD-GPR-DeepAR, featuring a unique data decomposition algorithm, monotone decomposition (MonoD). MonoD alleviates end effects by decoupling the original capacity signal into a smooth, decreasing trend and a fluctuant capacity regeneration term. Gaussian process regression (GPR) and deep autoregressive (DeepAR) models are then applied to the subseries for prediction, including uncertainty intervals. Validation using simulations and three real lithium-ion battery datasets demonstrates MonoD’s superior performance in capturing the authentic aging trajectory characteristics. Compared to alternative methods, the MonoD-GPR-DeepAR model shows its effectiveness in addressing complexities introduced by capacity regeneration phenomena in lithium-ion battery RUL prediction.

精准预测锂离子电池（lithium-ion batteries）的剩余使用寿命（remaining useful life, RUL）对于高效的设备健康管理至关重要。在电池老化过程中，其容量呈现非线性变化特征，且存在间歇性容量再生现象，会导致相邻循环周期间容量出现突增，这给建模与预测工作带来了显著挑战。尽管经验模态分解（empirical mode decomposition, EMD）常被用于分解电池容量序列，但绝大多数基于EMD的RUL预测方法仍存在诸多局限，包括端点效应、信息泄露问题，以及缺乏不确定性量化手段。为解决上述挑战，本文提出了一种新型RUL预测框架MonoD-GPR-DeepAR，其核心为一种独特的数据分解算法——单调分解（monotone decomposition, MonoD）。MonoD通过将原始容量信号解耦为平滑递减的趋势项与波动型容量再生项，有效缓解了端点效应问题。随后，研究人员将高斯过程回归（Gaussian process regression, GPR）与深度自回归（deep autoregressive, DeepAR）模型应用于各子序列开展预测，其中可输出不确定性区间。通过仿真实验与三个真实锂离子电池数据集开展验证，结果表明MonoD在捕捉真实老化轨迹特征方面性能更优。与其他替代方法相比，MonoD-GPR-DeepAR模型在应对锂离子电池RUL预测中由容量再生现象引发的复杂性问题时，展现出了良好的有效性。