Study of Derived Battery Features for Real-Time Estimation of SOH and RUL of EV Battery Using Data Analysis

2024-26-0123

01/16/2024

Features
Event
Symposium on International Automotive Technology
Authors Abstract
Content
EVs are extensively utilised with lithium-ion batteries. Predicting the SOH of batteries is desired to achieve optimal operation and health management. The most significant obstacle to accurately predicting battery health is choosing battery features. This study introduces numerous data analysis strategies to manage feature irrelevancy and help determine which features can be selected and used in real-time and edge computing. The first step in manually crafting features is to analyse the evolution pattern of numerous essential battery characteristics. Second, the correlation between selected features and degraded capacity was analysed. Then, selected features are fed into a representative machine learning regression model to effectively predict the remaining capacity of the battery to find the SOH status. Finally, the remaining battery capacity is selected as a feature to predict the RUL in terms of remaining charge-discharge cycles.
Meta TagsDetails
DOI
https://doi.org/10.4271/2024-26-0123
Pages
5
Citation
Nangare, K., Nidubrolu, K., and Gaikwad, P., "Study of Derived Battery Features for Real-Time Estimation of SOH and RUL of EV Battery Using Data Analysis," SAE Technical Paper 2024-26-0123, 2024, https://doi.org/10.4271/2024-26-0123.
Additional Details
Publisher
Published
Jan 16
Product Code
2024-26-0123
Content Type
Technical Paper
Language
English