A Long Short-Term Memory-Based Deep Learning Digital Twin of a Li-Ion Cell for Battery SOC Estimation

This study aims to implement the digital twin of a Li-ion battery by using real measurement data and to create a deep learning-based SOC (state of charge) estimation solution. In the case of the SOC estimator, a special type of deep learning, so-called long short-term memory (LSTM), was used to incr...

Full description

Saved in:
Bibliographic Details
Main Authors: József Richárd Lennert, Dénes Fodor
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Engineering Proceedings
Subjects:
Online Access:https://www.mdpi.com/2673-4591/79/1/16
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This study aims to implement the digital twin of a Li-ion battery by using real measurement data and to create a deep learning-based SOC (state of charge) estimation solution. In the case of the SOC estimator, a special type of deep learning, so-called long short-term memory (LSTM), was used to increase the capabilities of the estimator. The digital twin and the SOC estimator were created by using MATLAB and MATLAB/Simulink. As a result, the implemented system can accurately simulate the non-linearities of the Li-ion battery and provide a satisfactory estimation of the SOC of the battery.
ISSN:2673-4591