State of Charge Estimation of Lithium-Ion Batteries Based on Temporal Convolutional Network and Transfer Learning

Liu, Yuefeng; Li, Jiaqi; Zhang, Gong; Hua, Bin; Xiong, Naixue

doi:10.1109/access.2021.3057371

Cited by 88 publications

(34 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In general, SOC [ 3 , 17 , and 18 ] is the ratio of the remaining electric quantity to the rated electric quantity. Lithium-ion batteries have typical nonlinear characteristics, and it is difficult to measure the total power released by existing means or methods.…”

Section: Methodsmentioning

confidence: 99%

“…To simulate the engineering reality of these electric cars, there are generally FUDS(The Federal Urban Driving Schedule), DST(Dynamic Stress Test), US06(Urban Dynamometer Driving) Schedule, and BJDST(Beijing Dynamic Stress Test). In this paper, FUDS and BJDST [ 3 , 23 – 26 ] are used to test procedures through these two working conditions to test the multi-neural network training test.…”

Section: Methodsmentioning

confidence: 99%

“…In addition, the development of pure electric vehicles [ 2 ] is very rapid. The main reason is the energy data storage medium used in electric vehicles–lithium-ion battery [ 3 ]. Lithium-ion batteries have the advantages of long cycle life, high energy density, and low internal resistance.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

SOC Estimation of Lithium-Ion Battery for Electric Vehicle Based on Deep Multilayer Perceptron

Li¹,

Jiang²,

Guo

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

The state of charge (SOC) is one of the main indexes of the lithium-ion battery, which affects the practice range of new energy vehicles and the safety of the battery. Nevertheless, the value of SOC cannot be measured directly. At present, the algorithm for estimating the state of charge is not very satisfactory. The multilayer perceptron algorithm designed during this paper encompasses a sensible impact on state estimation. During this paper, the multilayer network is designed to estimate the charged state of lithium batteries from the three-layer artificial neural network to the eleven-layer artificial neural network. After preprocessing the dataset and comparing several activation functions, the ten-layer fully connected neural network is the most efficient to estimate the SOC. In order to prevent over-fitting of the multilayer perceptron algorithm, the two techniques of the BatchNormalization layer and Dropout layer work together to inhibit over-fitting. At the same time, the accuracy of extended Kalman filter, long and short memory network, and recurrent neural network are compared. The multilayer perceptron network designed during this paper has the highest accuracy. Finally, in the open dataset, both the training and test errors achieve good results. The algorithm developed in this paper has made some progress in SOC estimation.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

SOC Estimation of Lithium-Ion Battery for Electric Vehicle Based on Deep Multilayer Perceptron

Li¹,

Jiang²,

Guo

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

show abstract

“…At 0 C to 25 C, the method estimates an average MAE of 0.67%. 71 Table 3 provides the mean absolute error summary (MAE) based CNN.…”

Section: Convolutional Neural Networkmentioning

confidence: 99%

A comprehensive review on the state of charge estimation for lithium‐ion battery based on neural network

Cui

Wang

et al. 2021

Intl J of Energy Research

233

View full text Add to dashboard Cite

Implementing carbon neutrality and emission peak policies requires a highlevel electric vehicle field. Lithium-ion batteries have been considered an essential component of electric vehicle power batteries. Effective state of charge (SOC) estimation for lithium-ion batteries is a critical problem that needs to be addressed at present. With the feature extraction and fitting capability, the neural network can achieve accurate SOC estimation without considering the internal electrochemical state of the battery. This article overviews the definition of SOC and the relationship with battery aging state. Then, by examining recent literature on estimating the SOC of Lithium-ion batteries using neural network methods, the methods are classified into three categories: feed-forward neural network method, deep learning method, and hybrid method. The progress of neural network methods in SOC estimation applications is systematically reviewed, including principles, advantages, disadvantages, current status, and estimation errors. Possible recommendations for next-generation intelligent battery management systems and SOC estimation are also presented. This review's highlighted insights will inspire researchers in the battery field and point the way to developing electric vehicles.

show abstract

“…Where SOC0 is the initial state of charge of the battery tested, I (t) is the measured current, and Qnom is the nominal capacity of the battery. Furthermore, the voltage-current relationship is established through ( 4)- (6).…”

Section: B Equivalent Circuit Modelmentioning

confidence: 99%

A Novel Modal Representation of Battery Dynamics

et al. 2022

View full text Add to dashboard Cite

Electrochemical cells are complicated energy storage systems with nonlinear voltage dynamics.There is a need for accurate dynamic modeling of the battery system to predict its behavior over time when discharging. The study conducted in this paper developed an intuitive model for electrochemical cells based on a simple mechanical analogy. A three-degree-of-freedom, spring-mass-damper system is decomposed into modal coordinates that represent the overall discharge, mass transport, and double-layer effect of the electrochemical cell. The developed model was experimentally demonstrated through pulsed discharge tests of commercially available lithium-ion and nickel metal hydride cells. The modal parameters of the natural frequency and damping ratio for each mode were determined by numerically minimizing the error in the time responses. Additionally, the mechanical analog was applied to two datasets [23,24]. The first dataset was used to optimize the modal parameters whereas the second dataset was utilized to validate the tuned parameters. It was found that the modal representation of the mechanical analog could accurately predict the time-response dynamics of all the cells considered. Additionally, by considering the discharge modal coordinate, the open-circuit voltage was determined and validated to that measured experimentally from the voltage relaxation peaks.

show abstract

State of Charge Estimation of Lithium-Ion Batteries Based on Temporal Convolutional Network and Transfer Learning

Cited by 88 publications

References 33 publications

SOC Estimation of Lithium-Ion Battery for Electric Vehicle Based on Deep Multilayer Perceptron

SOC Estimation of Lithium-Ion Battery for Electric Vehicle Based on Deep Multilayer Perceptron

A comprehensive review on the state of charge estimation for lithium‐ion battery based on neural network

A Novel Modal Representation of Battery Dynamics

Contact Info

Product

Resources

About