Data Mining for Early Cycle Life Prediction in Lithium-Ion Battery Production

Sandro Stock; Mahmoud Ahmed; Fabian Konwitschny; Rudiger Daub

doi:10.1016/j.procir.2024.08.267

Data Mining for Early Cycle Life Prediction in Lithium-Ion Battery Production

Sandro Stock
, Mahmoud Ahmed
, Fabian Konwitschny
, Rudiger Daub

Chair of Production Technology and Energy Storage Systems

Technical University of Munich

Research output: Contribution to journal › Conference article › peer-review

1 Scopus citations

Abstract

The efficient prediction of product quality is a major challenge in lithium-ion battery production, as conventional measures such as aging are time-consuming and costly. This study presents a comprehensive data mining approach to predict the quality of lithium-ion batteries using linear and non-linear support vector machines. A methodology for extracting and selecting features from data sources within production is presented, and several feature selection algorithms - as well as models - are compared with regard to their predictive power. A minimum test error of 8.8 % for the early cycle life prediction was achieved, along with a classification accuracy of 96.6 %, when dividing the lithium-ion batteries into two quality grades with high and low cycle life.

Original language	English
Pages (from-to)	835-840
Number of pages	6
Journal	Procedia CIRP
Volume	126
DOIs	https://doi.org/10.1016/j.procir.2024.08.267
State	Published - 2024
Event	17th CIRP Conference on Intelligent Computation in Manufacturing Engineering, CIRP ICME 2023 - Naples, Italy Duration: 12 Jul 2023 → 14 Jul 2023

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

Battery production
Cycle life prediction
Incremental capacity analysis
Machine learning
Support vector regression

Access to Document

10.1016/j.procir.2024.08.267

Cite this

@article{4886091d38024501b005dcfdcac1930f,

title = "Data Mining for Early Cycle Life Prediction in Lithium-Ion Battery Production",

abstract = "The efficient prediction of product quality is a major challenge in lithium-ion battery production, as conventional measures such as aging are time-consuming and costly. This study presents a comprehensive data mining approach to predict the quality of lithium-ion batteries using linear and non-linear support vector machines. A methodology for extracting and selecting features from data sources within production is presented, and several feature selection algorithms - as well as models - are compared with regard to their predictive power. A minimum test error of 8.8 \% for the early cycle life prediction was achieved, along with a classification accuracy of 96.6 \%, when dividing the lithium-ion batteries into two quality grades with high and low cycle life.",

keywords = "Battery production, Cycle life prediction, Incremental capacity analysis, Machine learning, Support vector regression",

author = "Sandro Stock and Mahmoud Ahmed and Fabian Konwitschny and Rudiger Daub",

note = "Publisher Copyright: {\textcopyright} 2024 Elsevier B.V.. All rights reserved.; 17th CIRP Conference on Intelligent Computation in Manufacturing Engineering, CIRP ICME 2023 ; Conference date: 12-07-2023 Through 14-07-2023",

year = "2024",

doi = "10.1016/j.procir.2024.08.267",

language = "English",

volume = "126",

pages = "835--840",

journal = "Procedia CIRP",

issn = "2212-8271",

publisher = "Elsevier B.V.",

}

TY - JOUR

T1 - Data Mining for Early Cycle Life Prediction in Lithium-Ion Battery Production

AU - Stock, Sandro

AU - Ahmed, Mahmoud

AU - Konwitschny, Fabian

AU - Daub, Rudiger

PY - 2024

Y1 - 2024

N2 - The efficient prediction of product quality is a major challenge in lithium-ion battery production, as conventional measures such as aging are time-consuming and costly. This study presents a comprehensive data mining approach to predict the quality of lithium-ion batteries using linear and non-linear support vector machines. A methodology for extracting and selecting features from data sources within production is presented, and several feature selection algorithms - as well as models - are compared with regard to their predictive power. A minimum test error of 8.8 % for the early cycle life prediction was achieved, along with a classification accuracy of 96.6 %, when dividing the lithium-ion batteries into two quality grades with high and low cycle life.

AB - The efficient prediction of product quality is a major challenge in lithium-ion battery production, as conventional measures such as aging are time-consuming and costly. This study presents a comprehensive data mining approach to predict the quality of lithium-ion batteries using linear and non-linear support vector machines. A methodology for extracting and selecting features from data sources within production is presented, and several feature selection algorithms - as well as models - are compared with regard to their predictive power. A minimum test error of 8.8 % for the early cycle life prediction was achieved, along with a classification accuracy of 96.6 %, when dividing the lithium-ion batteries into two quality grades with high and low cycle life.

KW - Battery production

KW - Cycle life prediction

KW - Incremental capacity analysis

KW - Machine learning

KW - Support vector regression

UR - https://www.scopus.com/pages/publications/85208599354

U2 - 10.1016/j.procir.2024.08.267

DO - 10.1016/j.procir.2024.08.267

M3 - Conference article

AN - SCOPUS:85208599354

SN - 2212-8271

VL - 126

SP - 835

EP - 840

JO - Procedia CIRP

JF - Procedia CIRP

T2 - 17th CIRP Conference on Intelligent Computation in Manufacturing Engineering, CIRP ICME 2023

Y2 - 12 July 2023 through 14 July 2023

ER -

Data Mining for Early Cycle Life Prediction in Lithium-Ion Battery Production

Abstract

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this