Transfer Learning With Deep Neural Network for Capacity Prediction of Li-Ion Batteries Using EIS Measurement

Babaeiyazdi, Iman; Rezaei-Zare, Afshin; Shokrzadeh, Shahab

doi:10.1109/tte.2022.3170230

Cited by 17 publications

(4 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…TL involves transferring the acquired knowledge from a prior source dataset to facilitate the construction of models for a new target dataset. A minimal amount of freshly generated training data is sufficient to reconstruct an ML algorithm, even if the data do not stem from a similar test data distribution [194]. Transfer learning can be categorized into three types based on the changes that occur when moving from a source problem to a new target problem: inductive, transductive and unsupervised [195].…”

Section: ) Transfer Learningmentioning

confidence: 99%

“…Compared with the conventional data-driven method without TL, the proposed method reduces the error for the dT curve reconstruction by more than 20% and the SOH estimation error by more than 47% . In another study [194], TL in con-junction with a Deep Neural Network (DNN) was proposed for the capacity estimation of Li-ion batteries using EIS measurements as the input features to the base model. In this study, the base DNN model was trained and validated using a source dataset comprising EIS measurements and battery capacity at 25 o C and 35 o C. Then, followed by retraining and validation of the base model using the first 50% and first 20% of the target dataset at 45 o C. This created a new DNN-TL model that carried knowledge from the base model.…”

Section: ) Transfer Learningmentioning

confidence: 99%

“…TL has also been applied in health assessment in PHM FIGURE 19. DNN-TL with no fixed-layer and 1 fixed-layer compared with the stand-alone DNN model [194].…”

Section: ) Transfer Learningmentioning

confidence: 99%

See 2 more Smart Citations

Battery Reliability Assessment in Electric Vehicles: A State-of-the-Art

Omakor,

Miah,

Chaoui

2024

IEEE Access

View full text Add to dashboard Cite

Lithium-ion (Li-ion) batteries are being used in electric vehicles to reduce the reliance on fossil fuels due to their high energy density, design flexibility, and efficiency compared to other battery technologies. However, they undergo complex nonlinear degradation and performance declines when abused, making their reliability crucial for effective electric vehicle performance. This survey paper presents a comprehensive review of state-of-the-art battery reliability assessments for electric vehicles. First, the operating principle of Li-ion batteries, their degradation patterns, and degradation models are briefly discussed. Afterwards, the reliability assessments of Li-ion batteries are detailed in qualitative and quantitative approaches. The qualitative approach encompasses failure modes mechanisms and effects analysis, X-ray computed tomography, and scanning electron microscopy. In contrast, quantitative approaches involve multiphysics modelling, electrochemical impedance spectroscopy, incremental capacity and differential voltage analysis, machine learning, and transfer learning. Each technique is examined in terms of its principles, advantages, limitations, and applicability in Li-ion batteries for electric vehicles. Comparative analysis reveals that qualitative methods are primarily used in early design stages to assess potential risks and in post-mortem battery analysis in the laboratory, while quantitative techniques such as machine learning and transfer learning offer real-time prognostic health management and anomaly prevention. Also, the quantitative techniques tend to be more cost-effective compared to their counterparts. The potential for consolidating reliability methods through standardization of testing protocols, real-world data integration, controller area network use, and policy regulation are highlighted to guide further research.INDEX TERMS Capacity fade, causes of failure, data-driven, degradation trajectory, electric vehicle, electrochemical impedance spectroscopy, failure modes mechanisms and effects analysis, incremental capacity and differential voltage analysis, Li-ion batteries, machine learning, model-based, power fade, qualitative analysis, quantitative analysis, remaining useful life, reliability, state of health, state of charge, transfer learning, X-ray computed tomography.

show abstract

Section: ) Transfer Learningmentioning

confidence: 99%

Section: ) Transfer Learningmentioning

confidence: 99%

See 1 more Smart Citation

Battery Reliability Assessment in Electric Vehicles: A State-of-the-Art

Omakor,

Miah,

Chaoui

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The machine learning algorithms of the data-centric learning method assume that they are trained by the same distribution of train and test datasets [18]. However, in real-world applications, this assumption may not hold.…”

Section: Transfer Learning (Tl)mentioning

confidence: 99%

Transfer Learning with Deep Neural Network toward the Prediction of Wake Flow Characteristics of Containerships

Lee,

Lee

2023

JMSE

View full text Add to dashboard Cite

In this study, deep neural network (DNN) and transfer learning (TL) techniques were employed to predict the viscous resistance and wake distribution based on the positions of flow control fins (FCFs) applied to containerships of various sizes. Both methods utilized data collected through computational fluid dynamics (CFD) analysis. The position of the flow control fin (FCF) and hull form information were utilized as input data, and the output data included viscous resistance coefficients and components of propeller axial velocity. The base DNN model was trained and validated using a source dataset from a 1000 TEU containership. The grid search cross-validation technique was employed to optimize the hyperparameters of the base DNN model. Then, transfer learning was applied to predict the viscous resistance and wake distribution for containerships of varying sizes. To enhance the accuracy of feature prediction with a limited amount of data, learning rate optimization was conducted. Transfer learning involves retraining and reconfiguring the base DNN model, and the accuracy was verified based on the fine-tuning method of the learning model. The results of this study can provide hull designers for containerships with performance evaluation information by predicting wake distribution, without relying on CFD analysis.

show abstract

Interpretable machine learning prediction for li-ion battery's state of health based on electrochemical impedance spectroscopy and temporal features

Bao,

Liu,

et al. 2024

Electrochimica Acta

View full text Add to dashboard Cite

Transfer Learning With Deep Neural Network for Capacity Prediction of Li-Ion Batteries Using EIS Measurement

Cited by 17 publications

References 32 publications

Battery Reliability Assessment in Electric Vehicles: A State-of-the-Art

Battery Reliability Assessment in Electric Vehicles: A State-of-the-Art

Transfer Learning with Deep Neural Network toward the Prediction of Wake Flow Characteristics of Containerships

Interpretable machine learning prediction for li-ion battery's state of health based on electrochemical impedance spectroscopy and temporal features

Contact Info

Product

Resources

About