Li-ion Battery State of Charge Estimation Using Long Short-Term Memory Recurrent Neural Network with Transfer Learning

Vidal, Carlos; Kollmeyer, Phillip J.; Chemali, Ephrem; Emadi, Ali

doi:10.1109/itec.2019.8790543

Cited by 47 publications

(23 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [48], the authors have introduced a novel way to reduce training time and further improve SOC estimation by using an LSTM with Transfer learning, and in [49] the authors explored the accuracy impact of using different types of loss function optimizers during model training, e.g. Adam, NAdam, Adadelta, AdaGrad, RMSProp, and AdaMax.…”

Section: ) Gated Rnns Applied To Soc Estimationmentioning

confidence: 99%

Machine Learning Applied to Electrified Vehicle Battery State of Charge and State of Health Estimation: State-of-the-Art

et al. 2020

Self Cite

View full text Add to dashboard Cite

The growing interest and recent breakthroughs in artificial intelligence and machine learning (ML) have actively contributed to an increase in research and development of new methods to estimate the states of electrified vehicle batteries. Data-driven approaches, such as ML, are becoming more popular for estimating the state of charge (SOC) and state of health (SOH) due to greater availability of battery data and improved computing power capabilities. This paper provides a survey of battery state estimation methods based on ML approaches such as feedforward neural networks (FNNs), recurrent neural networks (RNNs), support vector machines (SVM), radial basis functions (RBF), and Hamming networks. Comparisons between methods are shown in terms of data quality, inputs and outputs, test conditions, battery types, and stated accuracy to give readers a bigger picture view of the ML landscape for SOC and SOH estimation. Additionally, to provide insight into how to best approach with the comparison of different neural network structures, an FNN and long short-term memory (LSTM) RNN are trained fifty times each for 3000 epochs. The error is somewhat different for each training repetition due to the random initial values of the trainable parameters, demonstrating that it is important to train networks multiple times to achieve the best result. Furthermore, it is recommended that when performing a comparison among estimation techniques such as those presented in this review paper, the compared networks should have a similar number of learnable parameters and be trained and tested with identical data. Otherwise, it is difficult to make a general conclusion regarding the quality of a given estimation technique. INDEX TERMS Machine learning, artificial intelligence, deep learning, battery management systems (BMS), electric vehicles, state of charge, state of health.

show abstract

Section: ) Gated Rnns Applied To Soc Estimationmentioning

confidence: 99%

Machine Learning Applied to Electrified Vehicle Battery State of Charge and State of Health Estimation: State-of-the-Art

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…With the development of artificial intelligence technology, applications in various fields have gradually matured [15][16][17][18]. Deep learning methods have also been introduced to the SOC estimation of lithium-ion batteries [19][20][21][22][23][24][25][26].…”

Section: Introductionmentioning

confidence: 99%

“…The data required by the training model must cover as much operation range as possible required by the application scene, and the trained model is only targeted at specific battery types. In the paper [22], the transfer learning technique was applied to the SOC estimation of lithium-ion battery. Based on the LSTM neural network, three kinds of models were implemented for SOC estimation for four different lithium-ion batteries, including models trained traditionally with no transfer learning, models trained with transfer learning using full target dataset, and models trained with transfer learning using partial target dataset.…”

Section: Introductionmentioning

confidence: 99%

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

Liu

Zhang

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Accurate estimation of the state of charge (SOC) is critical for the normal use of lithium-ion battery equipment like electric vehicles. However, the SOC of lithium-ion battery is not available by direct measure, but can only indirectly be estimated by measurable variables. According to the nonlinear characteristics between the measured values and SOC during the working period of lithium-ion batteries, we propose a method to estimate the SOC of lithium-ion batteries with Temporal Convolutional Network (TCN). The measured values of voltage, current, and temperature during the use of lithium-ion batteries can be directly mapped to accurate SOC in this method without using a battery model or adaptive filter. The network can self-learning and update parameters by being fed datasets collected under various working conditions and then obtain a model that can correctly estimate SOC under different estimation conditions. In addition, it can also be applied to different types of lithium-ion batteries through transfer learning with only a small amount of battery data. At various ambient temperature conditions, the average MAE estimated by the proposed method is 0.67% for all the tests, which proves that the TCN network is an effective tool to estimate the SOC of lithium-ion batteries.

show abstract

“…The data-driven model presented in this article combines LSTM-RNN architecture for the preliminary model and NARX-RNN architecture for the second-stage, finemodeling of the residuals of the LSTM-RNN model. Comparing with other methods described in the literature, 21,[46][47][48] hybrid LSTM-RNN/NARX-RNN model is a novel approach, requiring systematic research. Moreover, in the model's experimental evaluation, we compare the voltage estimation accuracy based on the data gathered for different SOC.…”

Section: Introductionmentioning

confidence: 99%

“…In Reference 47, there was the SOC and output voltage forecasting method of Li‐ion battery presented, using a hybrid of the VARMA (vector autoregressive moving average) and LSTM‐NN models for electric vehicle applications. In Reference 48, the SOC estimation of the four different Li‐ion battery types was evaluated using the deep neural network with LSTM layers, with transfer learning between models, showing that this approach reduces training time, and decreases the amount of required training data. In Reference 21, the authors compare the performance of the LSTM‐NN and a multilayer perceptron neural network (MLP‐NN) based models, for predicting SOC using predictors such as cell current, cell voltage, elapsed time, and cell temperature, where models were trained with the use of the MSE loss function, based on datasets from different charging/discharging rates.…”

Section: Introductionmentioning

confidence: 99%

Comparison of hybrid recurrent neural networks and dual‐polarization models of valve regulated lead acid battery

et al. 2020

View full text Add to dashboard Cite

Summary Distributed energy storage using low‐cost valve regulated lead acid (VRLA) batteries becomes a promising solution, especially in prosumer microinstallations. In herein work, the model‐based research of VRLA battery in pseudo‐random and hybrid power pulse cycles has been presented. The major goal of the article was to develop and compare parametric, analytical, and data‐driven models of the battery. There were chosen the dual polarization model and the hybrid recurrent neural network model based on long short‐term memory and nonlinear autoregressive with exogenous input model for estimating voltage values and determining nonmeasurable parameters of a VRLA battery for different values of the state of charge. Satisfactory results were obtained for both methods, with the favor of the dual polarization model. Finally, the advantages and disadvantages of both methods are discussed.

show abstract

Li-ion Battery State of Charge Estimation Using Long Short-Term Memory Recurrent Neural Network with Transfer Learning

Cited by 47 publications

References 5 publications

Machine Learning Applied to Electrified Vehicle Battery State of Charge and State of Health Estimation: State-of-the-Art

Machine Learning Applied to Electrified Vehicle Battery State of Charge and State of Health Estimation: State-of-the-Art

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

Comparison of hybrid recurrent neural networks and dual‐polarization models of valve regulated lead acid battery

Contact Info

Product

Resources

About