A Physics-Constrained Bayesian neural network for battery remaining useful life prediction

Najera-Flores, David; Hu, Zhen; Chadha, Mayank; Todd, Michael D.

doi:10.1016/j.apm.2023.05.038

Cited by 15 publications

(4 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Battery aging is primarily characterized by a decrease in available capacity and an increase in internal resistance, typically following a declining trajectory. To accurately describe the battery degradation trajectory, scholars have proposed various empirical models to describe the loss of battery capacity as a function of time or cycle numbers, including the linear model 48 , exponential model 49 , 50 , power-law model 51 , and failure forecast model (FFM) 52 , etc. These models all describe the battery’s degradation trajectory as a univariate function of time.…”

Section: Methodsmentioning

confidence: 99%

Physics-informed neural network for lithium-ion battery degradation stable modeling and prognosis

Wang,

Zhai,

Zhao

et al. 2024

Nat Commun

View full text Add to dashboard Cite

Accurate state-of-health (SOH) estimation is critical for reliable and safe operation of lithium-ion batteries. However, reliable and stable battery SOH estimation remains challenging due to diverse battery types and operating conditions. In this paper, we propose a physics-informed neural network (PINN) for accurate and stable estimation of battery SOH. Specifically, we model the attributes that affect the battery degradation from the perspective of empirical degradation and state space equations, and utilize neural networks to capture battery degradation dynamics. A general feature extraction method is designed to extract statistical features from a short period of data before the battery is fully charged, enabling our method applicable to different battery types and charge/discharge protocols. Additionally, we generate a comprehensive dataset consisting of 55 lithium-nickel-cobalt-manganese-oxide (NCM) batteries. Combined with three other datasets from different manufacturers, we use a total of 387 batteries with 310,705 samples to validate our method. The mean absolute percentage error (MAPE) is 0.87%. Our proposed PINN has demonstrated remarkable performance in regular experiments, small sample experiments, and transfer experiments when compared to alternative neural networks. This study highlights the promise of physics-informed machine learning for battery degradation modeling and SOH estimation.

show abstract

Section: Methodsmentioning

confidence: 99%

Physics-informed neural network for lithium-ion battery degradation stable modeling and prognosis

Wang,

Zhai,

Zhao

et al. 2024

Nat Commun

View full text Add to dashboard Cite

show abstract

“…On the other hand, ML can assist physicsbased models in sensitivity analysis, parameter identification, and quantifying the uncertainty of applied physical models. [197,198] Ding et al [199] provided a comprehensive review of related work in modeling battery degradation using electrochemical models, emphasizing the crucial role of ML-assisted model parameter sensitivity analysis and identification methods in understanding the entire parameter space. Streb et al [200] established an electrochemical model based on real-world driving data.…”

Section: Model Fusion and End-cloud Collaborationmentioning

confidence: 99%

Multiscale modeling for enhanced battery health analysis: Pathways to longevity

Yang,

Zhang,

Wang

et al. 2024

Carbon Neutralization

View full text Add to dashboard Cite

The issues of health assessment and lifespan prediction have always been prominent challenges in the large‐scale application of lithium‐ion batteries (LIBs). This paper reviews the multiscale modeling techniques and their applications in battery health analysis, including atomic scale computational chemistry, particle scale reaction simulations, electrode scale structural models, macroscale electrochemical models, and data‐driven models at the system level. Multiscale modeling offers a profound insight into material behavior and the aging process of batteries, thereby providing a valuable reference for both estimation and management strategies of battery state of health. To extend the battery lifespan, the utilization of artificial intelligence for material discovery and manufacturing process optimization, the implementation of end‐cloud collaborative battery management systems, and the design of a multiscale simulation integration platform are considered. A management framework aimed at extending battery life is further proposed. This framework offers a promising roadmap for addressing health analysis challenges in LIBs, ultimately leading to more reliable, efficient, and durable solutions for next‐generation batteries.

show abstract

“…This validated approach, used under various conditions, leverages an anode potential regulation, derived from a Newman-type P2D modeling framework, and showcasing a significant reduction in the risks of lithium plating. D. A. Najera-Flores et al [14] present a groundbreaking end-to-end deep learning framework for rapid lithium-ion battery RUL prediction. By emphasizing temporal patterns and cross-data correlations from raw data, like terminal voltage, current, and cell temperature, the approach achieves predictions 25X faster with a noteworthy 10.6% mean absolute error rate improvement.…”

Section: Related Workmentioning

confidence: 99%

Optimizing Electric Vehicle Battery Life: A Machine Learning Approach for Sustainable Transportation

Karthick,

Ravivarman,

Priyanka

2024

WEVJ

View full text Add to dashboard Cite

Electric vehicles (EVs) are becoming increasingly popular, due to their beneficial environmental effects and low operating costs. However, one of the main challenges with EVs is their short battery life. This study presents a comprehensive approach for predicting the Remaining Useful Life (RUL) of Nickel Manganese Cobalt-Lithium Cobalt Oxide (NMC-LCO) batteries. This research utilizes a dataset derived from the Hawaii Natural Energy Institute, encompassing 14 individual batteries subjected to over 1000 cycles under controlled conditions. A multi-step methodology is adopted, starting with data collection and preprocessing, followed by feature selection and outlier elimination. Machine learning models, including XGBoost, BaggingRegressor, LightGBM, CatBoost, and ExtraTreesRegressor, are employed to develop the RUL prediction model. Feature importance analysis aids in identifying critical parameters influencing battery health and lifespan. Statistical evaluations reveal no missing or duplicate data, and outlier removal enhances model accuracy. Notably, XGBoost emerged as the most effective algorithm, providing near-perfect predictions. This research underscores the significance of RUL prediction for enhancing battery lifecycle management, particularly in applications like electric vehicles, ensuring optimal resource utilization, cost efficiency, and environmental sustainability.

show abstract

A Physics-Constrained Bayesian neural network for battery remaining useful life prediction

Cited by 15 publications

References 53 publications

Physics-informed neural network for lithium-ion battery degradation stable modeling and prognosis

Physics-informed neural network for lithium-ion battery degradation stable modeling and prognosis

Multiscale modeling for enhanced battery health analysis: Pathways to longevity

Optimizing Electric Vehicle Battery Life: A Machine Learning Approach for Sustainable Transportation

Contact Info

Product

Resources

About