This paper introduces a novel two-stage identification algorithm for estimating the state of health (SOH) of batteries. The proposed approach leverages the unique structural properties of the SOH model by employing a sequential parameter estimation strategy. In the first stage, the parameters of the linear component are updated using a least squares algorithm. In the second stage, the nonlinear parameters are estimated via a gradient descent algorithm, building upon the previously identified linear parameters. This innovative methodology offers notable advantages over conventional gradient descent methods, including simpler implementation and improved convergence efficiency. The effectiveness of the algorithm is verified through comprehensive simulation studies, which confirm its superior performance characteristics.
WangSLFernandezCChenMJ, et al.A novel safety anticipation estimation method for the aerial lithium-ion battery pack based on the real-time detection and filtering. J Clean Prod2018; 185: 187–197.
2.
WangSLStroeDIFernandezC, et al.A novel power state evaluation method for the lithium battery packs based on the improved external measurable parameter coupling model. J Clean Prod2020; 242: 118506.
3.
WangSLGaoHYTakyi-AninakwaP, et al.Improved multiple feature-electrochemical thermal coupling modeling of lithium-ion batteries at low temperature with real-time coefficient correction. Prot Contr Mod Pow Syst2024; 9(3): 157–173.
4.
PattipatiBSankavaramCPattipatiK. System identification and estimation framework for pivotal automotive battery management system characteristics. IEEE Trans Syst Man Cybern C2011; 41(6): 869–884.
5.
WeiHZhongYFanL, et al.Design and validation of a battery management system for solar-assisted electric vehicles. J Power Sources2021; 513(11): 230531.
6.
NaseriFSchaltzEStroeDI, et al.An enhanced equivalent circuit model with real-time parameter identification for battery state-of-charge estimation. IEEE Trans Ind Electron2022; 69(4): 3743–3751.
7.
WangSLMaCGaoHY, et al.Improved hyperparameter Bayesian optimization-bidirectional long short-term memory optimization for high-precision battery state of charge estimation. Energy2025; 328: 136598.
8.
WangSLFanYCJinSY, et al.Improved anti-noise adaptive long short-term memory neural network modeling for the robust remaining useful life prediction of lithium-ion batteries. Reliab Eng Syst Saf2023; 230: 108920.
9.
ZhouYFWangSLXieYX, et al.Remaining useful life prediction and state of health diagnosis for lithium-ion batteries based on improved grey wolf optimization algorithm-deep extreme learning machine algorithm. Energy2023; 285: 128761.
10.
BhanguBSBentleyPStoneDA, et al.Nonlinear observers for predicting state-of-charge and state-of-health of lead-acid batteries for hybrid-electric vehicles. IEEE Trans Veh Technol2005; 54(3): 783–794.
11.
SangiriJBKulshreshthaTGhoshS, et al.A novel methodology to estimate the state-of-health and remaining-useful-life of a li-ion battery using discrete fourier transformation. J Energy Storage2022; 46(1): 103849.
12.
ChengLYChenJPuY, et al.Battery state of health model estimation based on variable projection algorithm using steffensen acceleration method. J Electrochem Soc2023; 170(5): 050509.
13.
ChengLYPuYChenJ, et al.Battery state of health estimation using variable projection algorithm based on truncated variable order fractional gradient descent. Ionics2024; 30: 863–874.
14.
LeiJXQuanXJFengS, et al.Accurate modeling of PLL with frequency-adaptive prefilter: on the positive feedback effect. IEEE Trans Power Electr2022; 37(4): 3747–3752.
15.
ZhangZLPuYXuF, et al.An improved adaptive Kalman filter based on auxiliary model for state of charge estimation with random missing outputs. J Electrochem Soc2023; 170: 020512.
16.
MartinezJARibaJRMoreno-EguilazM. State of health prediction of power connectors by analyzing the degradation trajectory of the electrical resistance. Mul Dig Pub Inst2021; 10(12): 1409–1424.
17.
ChenFWPadillaAPeterY, et al.Data-driven modeling of wireless power transfer systems with slowly time-varying parameters. IEEE Trans Power Electr2020; 35(11): 12442–12456.
18.
DingFXuLZhangX, et al.Hierarchical extended parameter identification methods and convergence for finite impulse response moving average models based on the hierarchical identification principle. Proc Inst Mech Eng Part I-J Syst Control Eng2025; 239: 1719–1746.
19.
ChenJZhuQMLiuYJ. Modified Kalman filtering based multi-step-length gradient iterative algorithm for ARX models with random missing outputs. Automatica2020; 118: 109034.
20.
WangJWXiongWLDingF. Identification of a non-commensurate fractional-order nonlinear system based on the separation scheme. Int J Adapt Control Signal Process2025; 39(1): 116–131.
21.
ChenJHuangBGanM, et al.A novel reduced-order algorithm for rational models based on Arnoldi process and Krylov subspace. Automatica2021; 129: 109663.
22.
WangJXiongWDingF. The Aitken Accelerated gradient algorithm for a class of dual-rate volterra nonlinear systems utilizing the self-organizing map technique. Int J Robust Nonlinear Control2025; 35(13): 5364–5379.
23.
ZhangZLChenJMaoYW, et al.Improved square root cubature Kalman filter for state of charge estimation with state vector outliers. Ionics2023; 29: 1369–1379.
24.
ChenJHuMFMaoYW, et al.Modified multi-direction iterative algorithm for separable nonlinear models with missing data. IEEE Sign Process Lett2022; 29: 1968–1972.
25.
ChenJMaJXGanM, et al.Multidirection gradient iterative algorithm: a unified framework for gradient iterative and least squares algorithms. IEEE Trans Automat Contr2022; 67(12): 6770–6777.
26.
XuLZhangXDingF, et al.A delta operator state estimation algorithm for discrete-time systems with state time-delay. IEEE Signal Process Lett2025; 32: 391–395.
27.
SahaBGoebelKPollS, et al.Prognostics methods for battery health monitoring using a bayesian framework. IEEE Trans Instrum Meas2009; 58(2): 291–296.
28.
GolubGPereyraV. Separable nonlinear least squares: the variable projection method and its applications. Speech Commun2003; 45(1): 63–87.
29.
ChenGYGanMChenC, et al.A two-stage estimation algorithm based on variable projection method for GPS positioning. IEEE Trans Instrum Meas2018; 67(11): 2518–2525.
30.
GanMSuXChenG, et al.Online learning under a separable stochastic approximation framework. IEEE Trans Pattern Anal Mach Intell2025; 47(2): 1317–1330.
31.
ChenGXuePGanM, et al.Variable projection algorithms: theoretical insights and a novel approach for problems with large residual. Automatica2025; 177: 112300.
32.
DingFXuLZhangX, et al.Recursive identification methods for general stochastic systems with colored noises by using the hierarchical identification principle and the filtering identification idea. Annu Rev Control2024; 57: 100942.
33.
WangWLiuSJiangY, et al.A robust filter and smoother-based expectation–maximization algorithm for bilinear systems with heavy-tailed noise. Mech Syst Signal Process2025; 236: 112912.
34.
JiYLiuJLiuH. An identification algorithm of generalized time-varying systems based on the taylor series expansion and applied to a pH process. J Process Control2023; 128: 103007.
35.
XuLDingFZhuQ. Separable synchronous multi-innovation gradient-based iterative signal modeling from on-line measurements. IEEE Trans Instrum Meas2022; 71: 1–13.
36.
JiYJiangAN. Filtering-based accelerated estimation approach for generalized time-varying systems with disturbances and colored noises. IEEE Trans Circuits Syst II2023; 70(1): 206–210.
37.
DingFXuLLiuP, et al.Two-stage parameter estimation methods for linear time-invariant continuous-time systems. Syst Control Lett2025; 204: 106166.
38.
WangJWJiYZhangX, et al.Two-stage gradient-based iterative algorithms for the fractional-order nonlinear systems by using the hierarchical identification principle. Int J Adapt Control Signal Process2022; 36(7): 1778–1796.
39.
DingFXuLZhangX, et al.Hierarchical generalized extended parameter identification for multivariable equation-error ARMA-like systems by using the filtering identification idea. Annu Rev Control2025; 60: 100993.
40.
ZouYBXuL. Parameter identification and open-circuit voltage estimation of lithium-ion battery circuit model based on multi-innovation theory. Circuit Theory & Apps. 2025; cta.70128.
41.
NiJYLiuSY. Multi-innovation recursive methods for a class of nonlinear time series models based on the penalty term. Int J Adapt Control Signal Process. 2025. doi:10.1002/acs.4062.
42.
DingFLuanXXuL, et al.Hierarchical recursive gradient parameter identification for multi‐input ARX systems with partially‐coupled information vectors. Int J Adapt Control Signal Process2025; 39: 1978–1995.
43.
JiangMWuQLiX. Multisource heterogeneous data fusion analysis of regional digital construction based on machine learning. J Sens2022; 2022: 1–11.
44.
ZuoXToamQZhongY. Comparative analysis of hydrogen production methods: environmental impact and efficiency of electrochemical and thermochemical processes. Int J Hydrogen Energy2025; 118: 426–440.
45.
WangGHuangYZhangY, et al.Accurate identification of anxiety and depression based on the dlPFC in an emotional autobiographical memory task: a machine learning approach. Biomed Signal Process Control2025; 104: 107503.
46.
ZhouHWeiCHeJ, et al.Spatiotemporal soliton solutions in three dimensional combined linear-harmonic potentials with varying sources. Opt Quantum Electron2024; 56(8): 1338.
47.
HuangZHuWZhuZ, et al.TMSF: Taylor expansion approximation network with multi-stage feature representation for optical flow estimation. Digit Signal Process2025; 162: 105157.
48.
DengYMaJWuZ, et al.DSR-Net: distinct selective rollback queries for road cracks detection with detection transformer. Digit Signal Process2025; 164: 105266.
49.
ZhuZXiaMXuB, et al.GTEA: guided Taylor expansion approximation network for optical flow estimation. IEEE Sensors J2024; 24(4): 5053–5061.
50.
ZhuZHuangCXiaM, et al.RFRFlow: recurrent feature refinement network for optical flow estimation. IEEE Sensors J2023; 23(21): 26357–26365.
51.
DingF. System identification-performance analysis for identification methods. Beijing: Science Press, 2014.
52.
LiuYHuangGLiJ, et al.MA-GRNN: a high-efficient modeling attack approach utilizing generalized regression neural network for XOR arbiter physical unclonable functions. IEICE Electron Express2023; 20(13): 20230141.
53.
ZhangPHuCZhouQ, et al.Nonlinear control for soliton interactions in optical fiber systems. Nonlinear Dyn2020; 101(2): 1215–1220.
54.
LianHLiSJiangS, et al. Sensitivity analysis of space parameters for integrated natural circulation reactor under the rolling condition coupling power increase. Nucl Technol. 2025; 1–14.
55.
LiuJFuYHabibiM, et al.Evaluation of mechanical behavior of textile microfibers. Acta Mech2025; 236(5): 3081–3094.
56.
JainDLiuXNeelakandanS, et al.Modeling of human action recognition using hyperparameter tuned deep learning model. J Electron Imaging2023; 32(1): 011211.
57.
HuangZLinCXuB, et al.T2EA: target-aware Taylor expansion approximation network for infrared and visible image fusion. IEEE Trans Circuits Syst Video Technol2025; 35(5): 4831–4845.
58.
ZhuZAnQHuangC, et al.PDTE: pyramidal deep Taylor expansion for optical flow estimation. Pattern Recognit Lett2024; 180: 107–112.
59.
HuangZLiQFangH, et al.Iterative weighted nuclear norm for X-ray cardiovascular angiogram image denoising. Signal Image Video Process2017; 11(8): 1445–1452.
60.
ChenJZhuQMHuMF, et al.Improved gradient descent algorithms for time-delay rational state-space systems: intelligent search method and momentum method. Nonl Dynam2020; 101: 361–373.
61.
NesterovYE. A method of solving a convex programming problem with convergence rate o(1/k2). Dokl Akad Nauk SSSR1983; 269(3): 543–547.
