State of energy estimation of lithium battery based on least squares and support vector regression

Abstract To enhance the estimation precision of lithium battery state of energy (SOE) and avoid the complex modeling and parameter identification process in the estimation process, a joint algorithm using the least square method (LS) and support vector regression (SVR) is proposed in this paper. Dat...

Full description

Saved in:
Bibliographic Details
Published inJournal of physics. Conference series Vol. 2656; no. 1; pp. 12023 - 12029
Main Authors Yao, Bowei, Liu, Gaoling, Liu, Baolei, Zhai, Kejiao, Wei, Jinyi, Yin, Xilin
Format Journal Article
LanguageEnglish
Published Bristol IOP Publishing 01.12.2023
Subjects
Algorithms
Least squares
Lithium batteries
Mathematical models
Parameter identification
Physics
Regression
State of charge
Support vector machines
Online AccessGet full text

Cover

More Information
Summary:Abstract To enhance the estimation precision of lithium battery state of energy (SOE) and avoid the complex modeling and parameter identification process in the estimation process, a joint algorithm using the least square method (LS) and support vector regression (SVR) is proposed in this paper. Data on the voltage, current, temperature, and charge state of the lithium battery are extracted and incorporated into the energy factor. The lithium battery SOE estimation model is designed by the least-squares support vector regression (LS-SVR) methods, while the grid optimization method is deployed to optimize the hyper-parameters. Finally, the SOE prediction of lithium batteries under a dynamic scenario is realized. Comparison results show that the energy factor chosen in our article accurately reflects the SOE of lithium batteries, and the MAE is within 1.5%.
ISSN:1742-6588
1742-6596
DOI:10.1088/1742-6596/2656/1/012023