State of health (SOH) estimation based on gated recurrent unit (GRU) neural network with attention mechanism for lithium-on battery

Öz

Lithium-ion batteries play an irreplaceable role in various areas of production and life as an efficient energy storage element. The state of health (SOH) of lithium-ion batteries is critical to the safe operation of the energy storage system. In this study, measurable data from the battery management system, such as voltage, current and temperature profiles that change due to aging, were used. Based on these data, the relationship between capacity and charging profiles is learned by neural networks. The experimental results achieved in this study are based on the NASA lithium-ion battery data set. It has been observed that compared to the LSTM, GRU and BiLSTM methods, which are deep learning methods, the proposed GRU method with attention mechanism is more successful in estimating the health of the battery in terms of average absolute percentage error by up to 35%, 27% and 20%, respectively. The simulation studies were carried out using the deep learning toolbox in the MATLAB environment. In recent years, attention mechanisms have emerged as a powerful tool to improve the performance of time series forecasting models. In this study, LSTM, BiLSTM and GRU, which are used to solve time series problems, were tested on the same NASA data sets, and GRU was preferred because it is faster and simpler than these three methods. The mechanism proposed in this study is the SoH prediction mechanism created by combining GRU and Attention Mechanism.

Anahtar Kelimeler

• Lithium-ion battery
• neural network
• remaining useful life (RUL)
• capacity estimation
• battery state of health (SoH)
• LSTM
• GRU
• Bi-directional LSTM
• Self-Attention Mechanism

Lityum iyon batarya için dikkat mekanizmalı geçitli tekrarlayan birim (GRU) sinir ağını temel alan sağlık durumu (SOH) tahmini

Öz

Günümüzde lityum iyon bataryalar, verimli bir enerji depolama elemanı olarak üretimin ve yaşamın çeşitli alanlarında yeri doldurulamaz bir rol oynamaktadır. Lityum iyon bataryaların sağlık durumu (SOH), enerji depolama sisteminin güvenli çalışması için kritik öneme sahiptir. Bu çalışmada batarya yönetim sisteminden gelen, yaşlanmaya bağlı olarak değişen gerilim, akım ve sıcaklık profilleri gibi ölçülebilir veriler kullanılmıştır. Bu verilere dayanarak kapasite ile şarj profilleri arasındaki ilişki sinir ağları tarafından öğrenilir. Bu çalışmada ulaşılan deneysel sonuçlar NASA lityum iyon pil veri setine dayanmaktadır. Önerilen dikkat mekanizmalı GRU yöntemi, bataryanın sağlığının tahmininde ortalama mutlak yüzde hata açısından derin öğrenme yöntemlerinden olan LSTM, GRU ve BiLSTM yöntemlerine kıyasla sırasıyla %35, %27 ve %20’ye kadar daha başarılı olduğu görülmüştür. Yapılan benzetim çalışmaları MATLAB ortamında derin öğrenme toolbox’ı kullanılarak gerçekleştirilmiştir. Son yıllarda dikkat mekanizmaları, zaman serisi tahmin modellerinin performansını artırmak için güçlü bir araç olarak ortaya çıkmıştır. Bu çalışmada, zaman serisi problemlerinin çözümlerinde kullanılan LSTM, BiLSTM ve GRU aynı NASA veri setleri üzerinde denenmiş ve bu üç yöntemden daha hızlı ve basit olmasıyla GRU tercih edilmiştir. Bu çalışmada önerilen mekanizma GRU ile Dikkat Mekanizmasını birleştirerek oluşturulmuş SoH öngörüm mekanizmasıdır.

Anahtar Kelimeler

• Lityum iyon pil
• sinir ağı
• kalan kullanım ömrü (RUL)
• kapasite tahmini
• pilin sağlık durumu (SoH)
• LSTM
• GRU
• Çift Yönlü LSTM
• Dikkat Mekanizması

Kaynakça

1. L. Zhang, T. Ji, S. Yu, and G. Liu, Accurate prediction approach of soh for lithium-ion batteries based on LSTM method. Batteries, 9 (3), 177, 2023. https://doi .org/10.3390/batteries9030177
2. E. İ. Tezde and H. İ. Okumuş, Batarya Modelleri ve Şarj Durumu (SoC) Belirleme. EMO Bilimsel Dergisi, 8 (1), 21–25, 2018.
3. E. Zio, Prognostics and Health Management (PHM): Where are we and where do we (need to) go in theory and Practice. Reliability Engineering; System Safety, 218, 108119, 2022. http://dx.doi.org/10.1016/j.ress.20 21.108119
4. ISO 13381-1:2004, Condition Monitoring and Diagnostics of Machines Prognostics Part1: General Guidelines. International Standards Organisation, Geneva, Switzerland, Nov. 2004.
5. T. Sarıkurt and A. Balıkçı, A relative capacity estimation method for lithium batteries used in electric vehicle applications. Pamukkale University Journal of Engineering Sciences, 24 (5), 809–816, 2018. http://doi:10.5505/pajes.2018.58224
6. E. Çavuş and İ. Sancaktar, Batarya sağlık durumunun makine öğrenmesi ile kestirimi. NOHU J. Eng. Sci., 11 (3), 601–610, 2022, https://doi.org/10.28948/ngum uh.1112985
7. Y. Jiang, Y. Chen, F. Yang, and W. Peng, State of health estimation of lithium-ion battery with automatic feature extraction and self-attention learning mechanism. Journal of Power Sources, 556, 232466, 2023. https://doi.org/10.1016/j.est.2023.109690
8. Time Series forecasting methods, Techniques & Models, InfluxData, https://www.influxdata.com/time-series-forecasting-methods/, Accessed 3 December 2023.

9. Y. Choi, S. Ryu, K. Park, and H. Kim, Machine learning-based lithium-ion battery capacity estimation exploiting multi-channel charging profiles. IEEE Access, 7, 75143–75152, 2019. https://doi.org/10.1002 /er.7160
10. I. Jorge, T. Mesbahi, A. Samet, and R. Boné, Time series feature extraction for lithium-ion batteries state-of-health prediction. Journal of Energy Storage, 59, 106436, 2023. https://doi.org/10.1016/ j.est.2022.1064 36
11. J. Zhang, J. Hou, and Z. Zhang, Online state-of-health estimation for the lithium-ion battery based on an LSTM neural network with attention mechanism. 2020 Chinese Control and Decision Conference (CCDC), 2020. https://doi:10.1109/ccdc49329.2020.9164547
12. Y. Guo, D. Yang, K. Zhao, and K. Wang, State of Health Estimation for lithium-ion battery based on bi-directional long short-term memory neural network and attention mechanism. Energy Reports, 8, 208–215, 2022. https://doi:10.1016/j.egyr.2022.10.128
13. Gated Recurrent Unit Networks. https://www.geeks forgeeks.org/gated-recurrent-unit-networks/amp, Accessed 3 December 2023.
14. Prognostics Center of Excellence Data Set Repository, NASA. https://www.nasa.gov/intelligent-systems-divi sion/discovery-and-systems-health/pcoe/pcoe-data-set-repository/, Accessed 3 December 2023.
15. L. He et al., Battery-aware Mobile Data Service. IEEE Transactions on Mobile Computing, 16 (6), 1544–1558, 2017. http://doi:10.1109/tmc.2016.2597842
16. M. Rashid and A. Gupta, Effect of relaxation periods over cycling performance of a Li-Ion Battery. Journal of The Electrochemical Society, 162 (2), 2015. https://doi:10.1149/2.0201502jes
17. J. Wu, C. Zhang, and Z. Chen, An online method for lithium-ion battery remaining useful life estimation using importance sampling and neural networks. Applied Energy, 173, 134–140, 2016. https://doi. org/10.1016/j.apenergy.2016.04.057
18. L. Yang, G. Liu, Y. Dai, J. Wang, and J. Zhai, Detecting stealthy domain generation algorithms using heterogeneous deep neural network framework. IEEE Access, 8, 82876–82889, 2020. https://doi:10.1109/ access.2020.2988877