Dataset Information

Data-driven prediction of battery failure for electric vehicles.

ABSTRACT: Despite great progress in battery safety modeling, accurately predicting the evolution of multiphysics systems is extremely challenging. The question on how to ensure safety of billions of automotive batteries during their lifetime remains unanswered. In this study, we overcome the challenge by developing machine learning techniques based on the recorded data that are uploaded to the cloud. Using charging voltage and temperature curves from early cycles that are yet to exhibit symptoms of battery failure, we apply data-driven models to both predict and classify the sample data by health condition based on the observational, empirical, physical, and statistical understanding of the multiscale systems. The best well-integrated machine learning models achieve a verified classification accuracy of 96.3% (exhibiting an increase of 20.4% from initial model) and an average misclassification test error of 7.7%. Our findings highlight the need for cloud-based artificial intelligence technology tailored to robustly and accurately predict battery failure in real-world applications.

SUBMITTER: Zhao J

PROVIDER: S-EPMC9010759 | biostudies-literature | 2022 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Data-driven prediction of battery failure for electric vehicles.

Zhao Jingyuan J Ling Heping H Wang Junbin J Burke Andrew F AF Lian Yubo Y

iScience 20220328 4

Despite great progress in battery safety modeling, accurately predicting the evolution of multiphysics systems is extremely challenging. The question on how to ensure safety of billions of automotive batteries during their lifetime remains unanswered. In this study, we overcome the challenge by developing machine learning techniques based on the recorded data that are uploaded to the cloud. Using charging voltage and temperature curves from early cycles that are yet to exhibit symptoms of batter ...[more]

PMID: 35434566

Similar Datasets

Project description:Background: The correct design of electric vehicle (EV) charging infrastructures is of fundamental importance to maximize the benefits for users and infrastructure managers. In addition, the analysis and management of recharges can help evaluate integration with auxiliary systems, such as renewable energy resources and storage systems. EV charging data analysis can highlight informative behaviours and patterns for charging infrastructure planning and management. Methods: We present the analysis of two datasets about the recorded energy and duration required to charge EVs in the cities of Barcelona (Spain) and Turku (Finland). In particular, we investigated hourly, daily and seasonal patterns in charge duration and energy delivered. Simulated scenarios for the power request at charging stations (CSs) were obtained using statistical parameters of the Barcelona dataset and non-parametric distributions of the arrivals. Monte Carlo simulations were used to test different scenarios of users' influx at the CSs, and determine the optimal size of an integrated renewable energy system (RES). Results: This study highlighted the difference between fast and slow charging users' habits by analysing the occupancy at the charging stations. Aside from the charge duration, which was shorter for fast charges, distinct features emerged in the hourly distribution of the requests depending on whether slow or fast charges are considered. The distributions were different in the two analysed datasets. The investigation of CS power fluxes showed that results for the investment on a RES could substantially vary when considering synthetic input load profiles obtained with different approaches. The influence of incentives on the initial RES cost were investigated. Conclusions: The novelty of this work lies in testing the impact of different approach to design synthetic profiles in the determination of the optimal size of a photovoltaic (PV) system installed at a charging infrastructure, using the economic criterion of the net present value (NPV).

Dataset Information

Data-driven prediction of battery failure for electric vehicles.

Publications

Data-driven prediction of battery failure for electric vehicles.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets