Cylindrical Battery Fault Detection Under Extreme Fast Charging: A Physics-Based Learning Approach

Firoozi, Roya; Sattarzadeh, Sara; Dey, Satadru

doi:10.1109/tec.2021.3112950

Cited by 23 publications

(2 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, battery thermal fault diagnostics have received considerable awareness as many model-based approaches [13][14][15][16][17][18][19][20] along with data-driven and signal processing techniques [21][22][23] have been proposed. The review papers [24][25][26] provide a comprehensive list of existing approaches.…”

Section: Literature Review and Research Gapsmentioning

confidence: 99%

Thermal Fault Detection and Localization Framework for Large Format Batteries

Sattarzadeh,

Roy,

Dey

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Safety against thermal failures is crucial in battery systems. Real-time thermal diagnostics can be a key enabler of such safer batteries. Thermal fault diagnostics in large format pouch or prismatic cells pose additional challenges compared to cylindrical cells. These challenges arise from the fact that the temperature distribution in large format cells is at least two-dimensional in nature (along length and breadth) while such distribution can be reasonably approximated in one dimension (along radial direction) in cylindrical cells. This difference makes the placement of temperature sensor(s) non-trivial and the design of detection algorithm challenging. In this work, we address these issues by proposing a framework that (i) optimizes the sensor locations to improve detectability and isolability of thermal faults, and (ii) designs a filtering scheme for fault detection and localization based on a two dimensional thermal model. The proposed framework is illustrated by experimental and simulation studies on a commercial battery cell.

show abstract

Section: Literature Review and Research Gapsmentioning

confidence: 99%

Thermal Fault Detection and Localization Framework for Large Format Batteries

Sattarzadeh,

Roy,

Dey

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, existing methods often fail to adequately consider the non-stationarity and complex dynamics of battery data. Physical models, which are constructed based on the chemical and physical laws of batteries, theoretically provide a basis for predicting and interpreting battery states [13], [14]. Nevertheless, these models frequently demonstrate their limitations when confronted with the complexity of battery behavior in real-world applications.…”

Section: Introductionmentioning

confidence: 99%

Enhanced Wavelet Transform Dynamic Attention Transformer Model For Lithium-ion Battery Anomaly Detection

Liu,

Huang

2024

Preprint

View full text Add to dashboard Cite

In the context of rapid advancements in electric vehicle (EV) technology, the safety and reliability of lithium-ion (Li) batteries,as the core power source, have garnered significant attention. The non-stationarity and complex dynamics of battery data posemajor challenges in the field of battery state monitoring, making it difficult for existing monitoring and prediction methods toaccurately address the variability of real-world environments. To address these issues, we propose a Transformer model based ona wavelet transform dynamic attention mechanism (WADT), capable of effectively handling the non-stationarity of battery dataand capturing its complex dynamic changes. The dynamic attention mechanism, grounded in wavelet transform, can adaptivelyfocus on the most informative parts of the battery data, thereby enhancing the accuracy of anomaly detection. Building on thismechanism, we further developed a deep learning model that incorporates an improved Transformer architecture, specificallydesigned to handle the complex dynamics of time series data. This model not only considers the temporal dependencies of batterydata but also adapts to its non-stationary behavior, achieving more accurate and reliable anomaly detection. Experimental resultson public battery datasets have demonstrated the effectiveness of our proposed approach. Compared to existing technologies, ourmodel has achieved significant improvements in several key performance indicators, including but not limited to accuracy, AUCscore. These results not only validate the innovation and effectiveness of our method but also showcase its potential applicationin the field of battery anomaly detection.

show abstract