Joint estimation of State of Charge (SOC) and State of Health (SOH) for lithium ion batteries using Support Vector Machine (SVM), Convolutional Neural Network (CNN) and Long Sort Term Memory Network (LSTM) models

International Journal of Electrochemical Science, Journal Year: 2024, Volume and Issue: 19(9), P. 100747 - 100747

Published: July 31, 2024

This paper proposes a data-driven method for jointly estimating the State of Charge (SOC) and Health (SOH) batteries, addressing impact battery aging on SOC estimation. Initially, Support Vector Machine (SVM) is employed to estimate SOH battery, using constant voltage charging time discharging lithium-ion batteries as inputs, output. By training SVM model, accurate estimation achieved. Subsequently, rated capacity adjusted based estimated obtain current maximum available capacity. adjustment allows coupling SOC, resulting in that accounts factors. Leveraging advantages Convolutional Neural Networks (CNN) feature extraction Long Short-Term Memory (LSTM) neural networks handling long-term sequential data, CNN-LSTM model utilized The proposed utilizes Oxford Battery Dataset (Cells 1–8) NASA (B0005–B0007) estimation, (Cell 8) (B0007) results demonstrate Root Mean Square Error (RMSE) less than 0.81 % Absolute (MAE) 0.65 Cells 1–8, while B0005 B0007, RMSE 1.81 MAE 1.29 %. For show average over entire lifecycle Cell 8 0.3923 0.3339 %, whereas 0.6123 0.4976

Language: Английский

---

Economic implications of enhanced stock market analysis using improved forecasting methods

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127240 - 127240

Published: March 1, 2025

Language: Английский

---

Remaining Useful Life Prediction of Lithium-Ion Batteries Based on Neural Network and Adaptive Unscented Kalman Filter

Electronics, Journal Year: 2024, Volume and Issue: 13(13), P. 2619 - 2619

Published: July 4, 2024

Accurate prediction of remaining useful life (RUL) plays an important role in maintaining the safe and stable operation Lithium-ion battery management systems. Aiming at problem poor stability a single model, this paper combines advantages data-driven model-based methods proposes RUL method combining convolutional neural network (CNN), bi-directional long short-term memory (Bi-LSTM), SE attention mechanism (AM) adaptive unscented Kalman filter (AUKF). First, three types indirect features that are highly correlated with decay selected as inputs to model improve accuracy prediction. Second, CNN-BLSTM-AM is used further extract, select fuse form predictive measurements identified degradation metrics. In addition, we introduce AUKF increase uncertainty representation Finally, validated on NASA dataset CALCE compared other methods. The experimental results show able achieve accurate estimation RUL, minimum RMSE up 0.0030, MAE 0.0024, which has high robustness.

Language: Английский

---

Prediction of electric vehicle battery state of health estimation using a hybrid deep learning mechanism

International Journal of Green Energy, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 14

Published: Jan. 2, 2025

Lithium-ion batteries (LIBs) are widely employed, but fluctuations in temperature, overcharging, and overdischarging reduce their service lifetime. Battery health issues such as accelerated deterioration, loss of capacity, thermal runaway can also endanger battery safety functionality. This paper presents the integration a Bidirectional Recurrent Neural Network Long Short-Term Memory (biRNN-LSTM) network improve prediction capability Li-ion State Health (SoH) with complex patterns identification higher accuracy. Compared to traditional feed-forward neural networks, RNNs designed learn temporal dependencies perform sequence recognition on original data. After this, LSTM modules this by being an example long-term time series information, which helps solve problems vanishing gradients. To highlight effectiveness proposed method compare it Deep Convolutional (DCNN-LSTM), Gate Unit (GRU), (LSTM) from literature make accurate reliable predictions, Root Mean Square Error (RMSE), Maximum Accuracy (MAE), (MAX) assessment metrics were used for performance evaluation. GRU needs 8000 iterations identify SoH estimation errors because is less capable learning dependencies. The technique detect after 7000 since performs exceptionally well capturing fine-grained dynamics.

Language: Английский

---

State of health estimation for lithium-ion batteries based on optimal feature subset algorithm

Energy, Journal Year: 2025, Volume and Issue: unknown, P. 135685 - 135685

Published: March 1, 2025

Language: Английский

---

An improved co-training architecture for Lithium-ion batteries state of health estimation with semi-supervised learning

Journal of Power Sources, Journal Year: 2025, Volume and Issue: 643, P. 236928 - 236928

Published: April 17, 2025

Language: Английский

---

The Joint Estimation of SOC-SOH for Lithium-Ion Batteries Based on BiLSTM-SA

Electronics, Journal Year: 2024, Volume and Issue: 14(1), P. 97 - 97

Published: Dec. 29, 2024

Lithium-ion batteries are commonly employed in energy storage because of their extended service life and high density. This trend has coincided with the rapid growth renewable electric automobiles. However, as usage cycles increase, effectiveness diminishes over time, which can undermine both system’s performance security. Therefore, monitoring state charge (SOC) health (SOH) real time is particularly important. Traditional SOC calculation methods typically treat SOH independent variables, overlooking coupling between them. To tackle this issue, paper introduces a joint SOC-SOH estimation approach (BiLSTM-SA) that leverages bidirectional long short-term memory (BiLSTM) network combined self-attention (SA) mechanism. The proposed validated using publicly available dataset. With taken into account, MAE RMSE 0.84% 1.20%, showing notable increases accuracy relative to conventional methods. Additionally, it demonstrates strong robustness generalization across datasets multiple temperatures.

Language: Английский

---

An Analytical Approach for IGBT Life Prediction Using Successive Variational Mode Decomposition and Bidirectional Long Short-Term Memory Networks

Electronics, Journal Year: 2024, Volume and Issue: 13(20), P. 4002 - 4002

Published: Oct. 11, 2024

The precise estimation of the operational lifespan insulated gate bipolar transistors (IGBT) holds paramount significance for ensuring efficient and uncompromised safety industrial equipment. However, numerous methodologies models currently employed this purpose often fall short delivering highly accurate predictions. analytical approach that combines Pattern Optimization Algorithm (POA) with Successive Variational Mode Decomposition (SVMD) Bidirectional Long Short-term Memory (BiLSTM) network is introduced. Firstly, SVMD as an unsupervised feature learning method to partition data into intrinsic modal functions (IMFs), which are used eliminate noise preserve essential signal. Secondly, BiLSTM integrated supervised purposes, enabling prediction decomposed sequence. Additionally, hyperparameters penalty coefficients optimized utilizing POA technique. Subsequently, various predicted trained model, individual mode predictions subsequently aggregated yield model’s definitive final life prediction. Through case studies involving IGBT aging datasets, optimal model was formulated its capability validated. superiority proposed demonstrated by comparing it benchmark other state-of-the-art methods.

Language: Английский

---

Application of state of health estimation and remaining useful life prediction for lithium-ion batteries based on AT-CNN-BiLSTM

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Nov. 23, 2024

Ensuring the long-term safe usage of lithium-ion batteries hinges on accurately estimating State Health $$(\textrm{SOH})$$ and predicting Remaining Useful Life (RUL). This study proposes a novel prediction method based AT-CNN-BiLSTM architecture. Initially, key parameters such as voltage, current, temperature, SOH are extracted averaged for each cycle to ensure uniformity reliability input data. The CNN is utilized extract deep features from data, followed by BiLSTM analyze temporal dependencies in data sequences. Since multidimensional parameter used predict trend batteries, an attention mechanism employed enhance weight highly relevant vectors, improving model's analytical capabilities. Experimental results demonstrate that CNN-BiLSTM-Attention model achieves absolute error 0 RUL prediction, $$R^{2}$$ value greater than 0.9910 , MAPE less 0.9003 . Comparative analysis with hybrid neural network algorithms LSTM, BiLSTM, CNN-LSTM confirms proposed high accuracy stability estimation prediction.

Language: Английский

---

EnerNet: Attention-based dilated CNN-BILSTM for state of health prediction of CS2 prismatic cells in energy systems

Electrochimica Acta, Journal Year: 2024, Volume and Issue: 512, P. 145454 - 145454

Published: Dec. 7, 2024

Language: Английский

---