Co-estimation of parameters and state of charge for lithium-ion battery

•A recursive Bayesian algorithm is proposed for the online identification of battery parameters.•An online joint estimation algorithm of parameters and state of charge is proposed for the battery system.•Different initial SOC values and measurement noise covariance are applied to verify the robustne...

Full description

Saved in:
Bibliographic Details
Published in:Journal of electroanalytical chemistry (Lausanne, Switzerland) Vol. 907; p. 116011
Main Authors: Li, Junhong, Li, Lei, Li, Zheng, Jiang, Zeyu, Gu, Juping
Format: Journal Article
Language:English
Published: Amsterdam Elsevier B.V 15.02.2022
Elsevier Science Ltd
Subjects:
Accuracy
Adaptive algorithms
Adaptive extended Kalman filter
Algorithms
Dual polarization model
Dynamic characteristics
Electrical measurement
Extended Kalman filter
Lithium-ion batteries
Lithium-ion battery
Mathematical models
Model accuracy
Noise measurement
Parameter estimation
Parameter identification
Polarization
Product safety
Rechargeable batteries
Recursive Bayesian algorithm
Robustness
Service life
State of charge
Dual polarization model
Lithium-ion battery
Adaptive extended Kalman filter
State of charge
Recursive Bayesian algorithm
ISSN:1572-6657, 1873-2569
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•A recursive Bayesian algorithm is proposed for the online identification of battery parameters.•An online joint estimation algorithm of parameters and state of charge is proposed for the battery system.•Different initial SOC values and measurement noise covariance are applied to verify the robustness of the joint estimation method.•The proposed algorithm is validated under two dynamic current cycles. Accurate and reliable estimation of battery SOC is critical to enhance its service life and safety, and the accuracy of model parameter identification also directly affects the result of SOC estimation. This paper focuses on the parameter identification and SOC estimation of the dual-polarization. A co-estimation of recursive Bayesian algorithm and adaptive extended Kalman filter algorithm (RB-AEKF) is utilized to dynamically identify the model parameters and estimate the battery SOC. Firstly, the RB algorithm is used to identify model parameters through real-time current and voltage measurement data. The predicted voltage of the dual-polarization model is basically the same as the actual voltage, which better reflects the dynamic characteristics of the battery and the accuracy of the identification algorithm. In addition, adaptive noise variance updating algorithm is added to the extended Kalman filter to improve the accuracy of SOC estimation. Furthermore, a dataset consisting of data from a dynamic stress test (DST) and a federal urban driving schedule (FUDS) is used to verify the proposed method. The SOC estimafte error based on RB-AEKF stays within 2% and root mean square error is 0.01085 under FUDS test. Finally, we conduct the robustness analysis, and the results show that the algorithm has satisfactory robustness against inaccurate initial SOC and different measurement noise covariance.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1572-6657
1873-2569
DOI:10.1016/j.jelechem.2022.116011