Rapid failure mode classification and quantification in batteries: A deep learning modeling framework

Kim, Sangwook; Yi, Zonggen; Chen, Bor-Rong; Tanim, Tanvir R.; Dufek, Eric J.

doi:10.1016/j.ensm.2021.07.016

Cited by 44 publications

(24 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, as the model cannot simulate polarization, only OCV data were generated with preset aging mechanisms and degradation rates. Therefore, aging diagnosis methods based on such synthetic datasets are limited to OCV based inputs 48,49 …”

Section: Prediction Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Data‐driven battery degradation prediction: Forecasting voltage‐capacity curves using one‐cycle data

Tian

Xiong

Shen

et al. 2022

EcoMat

View full text Add to dashboard Cite

With the wide deployment of rechargeable batteries, battery degradation prediction has emerged as a challenging issue. However, battery life defined by capacity loss provides limited information regarding battery degradation. In this article, we explore the prediction of voltage‐capacity curves over battery lifetime based on a sequence to sequence (seq2seq) model. We demonstrate that the data of one present voltage‐capacity curve can be used as the input of the seq2seq model to accurately predict the voltage‐capacity curves at 100, 200, and 300 cycles ahead. This offers an opportunity to update battery management strategies in response to the predicted consequences. Besides, the model avoids feature engineering and is flexible to incorporate different numbers of input and output cycles. Therefore, it can be easily transplanted to other battery systems or electrochemical components. Furthermore, the model features data generation, that is, we can use the data of only one cycle to generate a large spectrum of aging data at the future cycles for developing other battery diagnosis or prognosis methods. In this way, the time and energy consuming battery degradation tests can be sharply reduced.

show abstract

Section: Prediction Resultsmentioning

confidence: 99%

“…An attractive alternative is synthetic datasets. Li et al 47 48,49 In this study, we explore the ability of the developed model to generate voltage-capacity curves. The generated data will be promising to be used to assist with the development of machine learning models for battery management.…”

Section: Data Generation Using One Cyclementioning

confidence: 99%

Data‐driven battery degradation prediction: Forecasting voltage‐capacity curves using one‐cycle data

Tian

Xiong

Shen

et al. 2022

EcoMat

View full text Add to dashboard Cite

show abstract

“…Recently, Kim et al developed a synthetic IC model-based deep learning rapid failure mode classification and quantification framework, automatizing aging more classification and quantification at the same time using significantly less amount of data. [132] Other in situ/operando cell-level sophisticated measurements include spatially resolved high-power (synchrotron) X-ray diffraction techniques that can not only provide insights into aging modes but also the underlying mechanisms at local or global scales. [44,116]…”

Section: Cell-level Measurementsmentioning

confidence: 99%

Enabling Extreme Fast‐Charging: Challenges at the Cathode and Mitigation Strategies

Tanim

Weddle

Yang

et al. 2022

Advanced Energy Materials

Self Cite

View full text Add to dashboard Cite

Charging lithium‐ion batteries (LiBs) in 10 to 15 min via extreme fast‐charging (XFC) is important for the widespread adoption of electric vehicles (EVs). Lately, the battery research community has focused on identifying XFC bottlenecks and determining novel design solutions. Like other LiB components, cathodes can present XFC bottlenecks, especially when considering long‐term battery life. Therefore, it is necessary to develop a comprehensive understanding of how XFC conditions degrade LiB cathodes. The present article reviews relevant cathode‐focused studies and summarizes the current understanding regarding cathode performance and aging issues under XFC conditions. Dominant aging modes and mechanisms are identified at different length‐scales with electrochemical correlations for LiNixMnyCozO2 (NMC)‐based cathodes. A range of electrochemical techniques and models provide key insights into cathode performance and life issues. A suite of multimodal and multiscale microscopy and X‐ray techniques is surveyed to quantify chemical, structural, and crystallographic NMC‐cathode degradation. Cathode cycle‐life is scaled to equivalent EV miles to illustrate how cathode degradation translates to real‐world scenarios and quantifies cathode‐related bottlenecks that hinder XFC adoption. Finally, the article discusses several cathode cycle‐life aging mitigation strategies with example case studies and identifies remaining challenges.

show abstract

“…Generation of synthetic data sets using physics-based models may be well suited for training generalizable models reliant on electrochemical signals to identify knee pathways or predict knees. 43,172 Conclusions and future work…”

Section: Modeling and Prediction Outlookmentioning

confidence: 99%

"Knees" in lithium-ion battery aging trajectories

Attia,

Bills,

Planella

et al. 2022

Preprint

View full text Add to dashboard Cite

Lithium-ion batteries can last many years but sometimes exhibit rapid, nonlinear degradation that severely limits battery lifetime. In this work, we review prior work on "knees" in lithium-ion battery aging trajectories. We first review definitions for knees and three classes of "internal state trajectories" (termed snowball, hidden, and threshold trajectories) that can cause a knee. We then discuss six knee "pathways", including lithium plating, electrode saturation, resistance growth, electrolyte and additive depletion, percolation-limited connectivity, and mechanical deformation-some of which have internal state trajectories with signals that are electrochemically undetectable. We also identify key design and usage sensitivities for knees. Finally, we discuss challenges and opportunities for knee modeling and prediction. Our findings illustrate the complexity and subtlety of lithium-ion battery degradation and can aid both academic and industrial efforts to improve battery lifetime.

show abstract

Rapid failure mode classification and quantification in batteries: A deep learning modeling framework

Cited by 44 publications

References 53 publications

Data‐driven battery degradation prediction: Forecasting voltage‐capacity curves using one‐cycle data

Data‐driven battery degradation prediction: Forecasting voltage‐capacity curves using one‐cycle data

Enabling Extreme Fast‐Charging: Challenges at the Cathode and Mitigation Strategies

"Knees" in lithium-ion battery aging trajectories

Contact Info

Product

Resources

About