基于CPO-MSC-BiLSTM-Attention模型的单特征锂电池SOH估计研究

张 程; 邱荣伟

doi:10.19783/j.cnki.pspc.260144

欢迎来到能源电力期刊网！

English Version 期刊登录

您当前的位置：

首页 >

文章列表页 >

基于CPO-MSC-BiLSTM-Attention模型的单特征锂电池SOH估计研究

理论分析 | 更新时间：2026-09-23

- 基于CPO-MSC-BiLSTM-Attention模型的单特征锂电池SOH估计研究
- Research on SOH estimation for lithium-ion batteries based on a CPO-MSC-BiLSTM-Attention model with a single feature
- 电力系统保护与控制 2026年54卷第18期
- 作者机构：
- 作者简介：
- 基金信息：
- DOI：10.19783/j.cnki.pspc.260144
  中图分类号：
- 纸质出版：2026
- 稿件说明：
移动端阅览
张程, 邱荣伟. 基于CPO-MSC-BiLSTM-Attention模型的单特征锂电池SOH估计研究[J]. 电力系统保护与控制, 2026,54(18).

ZHANG Cheng, QIU Rongwei. Research on SOH estimation for lithium-ion batteries based on a CPO-MSC-BiLSTM-Attention model with a single feature[J]. Power System Protection and Control, 2026, 54(18).
张程, 邱荣伟. 基于CPO-MSC-BiLSTM-Attention模型的单特征锂电池SOH估计研究[J]. 电力系统保护与控制, 2026,54(18). DOI： 10.19783/j.cnki.pspc.260144.

ZHANG Cheng, QIU Rongwei. Research on SOH estimation for lithium-ion batteries based on a CPO-MSC-BiLSTM-Attention model with a single feature[J]. Power System Protection and Control, 2026, 54(18). DOI： 10.19783/j.cnki.pspc.260144.

摘要

针对低特征维度下锂离子电池健康状态(state of health

SOH)估计精度不足的问题，提出了一种仅以放电深度(depth of discharge

DOD)为输入特征的SOH估计方法。采用冠豪猪优化算法(crown porcupine optimization

CPO)对模型参数进行全局寻优，以提升预测精度与稳定性。模型结构引入多尺度卷积(multi-scale convolution

MSC)提取多时间尺度特征，结合双向长短期记忆网络(bidirectional long short-term memory

BiLSTM)增强时序建模能力，并通过注意力机制实现特征自适应加权。同时在卷积层后引入MixStyle层，以增强模型在不同工况和数据分布条件下的鲁棒性。实验结果表明，该模型在仅依赖DOD单一特征的情况下，仍可实现较高精度的SOH估计，并在泛化性能和鲁棒性方面优于对比方法。该方法为低特征维度下的电池SOH预测提供了一种有效途径。

Abstract

A state of health (SOH) estimation method using only depth of discharge (DOD) as the input feature is proposed to address the insufficient accuracy of lithium-ion batteries SOH estimation under low-dimensional feature conditions. The crown porcupine optimization (CPO) algorithm is employed for global parameter optimization to improve prediction accuracy and stability. The model introduces multi-scale convolution (MSC) to extract multi-time-scale features

and bidirectional long short-term memory (BiLSTM) networks are used to enhance temporal modeling capability. An attention mechanism is employed for adaptive feature weighting. In addition

a Mixstyle layer is incorporated after the convolution layers to improve robustness under different operating conditions and data distributions. Experimental results show that the proposed model achieves high-accuracy SOH estimation using only the single DOD feature and outperforms comparative methods in generalization performance and robustness. This method provides an effective approach for battery SOH estimation under low-dimensional feature conditions.

关键词

Keywords

references

KANT A, KUMAR M, SIHAG S. Prediction of electric vehicle battery state of health estimation using a hybrid deep learning mechanism[J]. International Journal of Green Energy, 2025: 1-14.

WU T, LIU S, WANG Z, et al. SOC and SOH joint estimation of lithium-ion battery based on improved particle filter algorithm[J]. Journal of Electrical Engineering Technology, 2022, 17(1): 307-317.

AKRAM S A, SOHAIB M, CHOI W. SOH estimation of lithium-ion batteries using distribution of relaxation times parameters and long short-term memory model[J]. Batteries, 2025, 11(5): 183-183.

TIAN A, CHEN Z, PAN Z, et al. Li-ion battery state of health estimation based on short random charging segment and improved long short-term memory[J]. IET Signal Processing, 2023.

HOU Z, XU W, JIA G, et al. Estimation of SOH of lithium-ion batteries based on PSO-Bi GRU-Attention network[J]. Journal of the Electrochemical Society, 2024, 171(2).

WEI J, WAN Y, ZHANG C, et al. Enhanced SOC and SOH estimation for Li-ion batteries based on combining adaptive central difference Kalman filter and discrete-time sliding mode observer[J]. Journal of Energy Storage, 2025, 114: 115671-115671.

NIANKAI Y, ZIYOU S, HEATH H, et al. Robust state of health estimation of lithium-ion batteries using convolutional neural network and random forest[J]. Journal of Energy Storage, 2022, 48.

ZHANG X, HOU J, WANG Z, et al. Joint SOH-SOC estimation model for lithium-ion batteries based on GWO-BP neural network[J]. Energies, 2022, 16(1): 132-132.

HU P, TANG F, LI H, et al. Joint state of charge (SOC) and state of health (SOH) estimation for lithium-ion batteries packs of electric vehicles based on NSSR- LSTM neural network[J]. Energies, 2023, 16(14).

FANG X, XU M, FAN Y. SOC-SOH estimation and balance control based on event-triggered distributed optimal Kalman consensus filter[J]. Energies, 2024, 17(3).

WEN J, ZOU Q, CHEN C, et al. Linear correlation between state-of-health and incremental state-of-charge in Li-ion batteries and its application to SoH evaluation[J]. Electrochimica Acta, 2022, 434.

REN J, MA J, WANG H, et al. A comprehensive review on research methods for lithium-ion battery of state of health estimation and end of life prediction: methods, properties, and prospects[J]. Protection and Control of Modern Power Systems, 2025, 10(3): 146-165.

ZENG M, ZHANG P, YANG Y, et al. SOC and SOH joint estimation of the power batteries based on fuzzy unscented Kalman filtering algorithm[J]. Energies, 2019, 12(16).

PARK J, LEE M, KIM G, et al. Integrated approach based on dual extended Kalman filter and multivariate autoregressive model for predicting battery capacity using health indicator and SOC/SOH[J]. Energies, 2020, 13(9).

RESHMA P, MANOHAR V J. Collaborative evaluation of SoC, SoP and SoH of lithium-ion battery in an electric bus through improved remora optimization algorithm and dual adaptive Kalman filtering algorithm[J]. Journal of Energy Storage, 2023, 68.

LI R, LIU P, LI K, et al. AdaBoost.Rt-LSTM based joint SOC and SOH estimation method for retired batteries[J]. Batteries, 2023, 9(8).

GUO Y, YANG D, ZHANG Y, et al. Online estimation of SOH for lithium-ion battery based on SSA-Elman neural network[J]. Protection and Control of Modern Power Systems, 2022, 79(3): 1-17.

SAXENA S, GOEBEL K. Battery data set[R]. NASA Ames Prognostics Data Repository, 2008.

PECHT M G, JAISWAL A. A prognostics and health management roadmap for information and electronics-rich systems[J]. Microelectronics Reliability, 2013, 53(3): 317-323.

ABDEL-BASSET M, MOHAMED R, ABOUHAWWASH M. Crested porcupine optimizer: a new nature-inspired metaheuristic[J]. Knowledge-Based Systems, 2024, 284.

ZHOU K, YANG Y, QIAO Y, et al. MixStyle neural networks for domain generalization and adaptation[J]. International Journal of Computer Vision, 2023, 132(3): 822-836.

WU L, CHEN C, LI Z, et al. The joint estimation of SOC-SOH for lithium-ion batteries based on BiLSTM- SA[J]. Electronics, 2024, 14(1): 97-97.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于轻量化特征筛选的高原牵引负荷超短期预测

基于时空卷积-注意力网络的分布式光伏缺失数据插补方法

基于深度残差网络和改进时序卷积神经网络的宽频振荡监测

基于Linear Transformer-CNN模型的继电保护系统缺陷文本深度挖掘技术

基于变分自编码器的多源数据融合窃电检测方法