Due to long cycle life and high power and energy density, Lithium-ion batteries have become the main energy storage device in several applications. Since a direct measurement of the battery State of Charge (SoC) is still difficult in many situations, a SoC accurate estimation is pivotal for controlling the battery status and ensuring its maintenance with the aim of reducing operational costs and preventing malfunctions and damages. In view of this, the aim of this paper was to estimate the SoC of rechargeable lithium polymer batteries during charge and discharge cycles by using two different Artificial Neural Networks (ANNs), i.e. Feedforward Neural Network (FNN) and Recurrent Neural Network (RNN). The ANNs were trained and tested with experimental charge and discharge cycles acquired with a home-made automated system. Then, the goodness of prediction of the two approaches was compared in terms of mean-squared error and relative error. Results demonstrated that for the charge cycle the SoC prediction was better with the Recurrent Neural Network approach, with a relative error lower than 5 %. On the other hand, for the discharge cycle the SoC prediction was quite similar between the FNN and RNN.

Apa, L., Del Prete, Z., Forconi, F., Palermo, M., Riganti fulginei, F., Rizzuto, E., et al. (2024). Neural Network Approaches for State of Charge Prediction of Rechargeable Lithium Polymer Batteries. In 2024 IEEE 22nd Mediterranean Electrotechnical Conference, MELECON 2024 (pp.236-241). Institute of Electrical and Electronics Engineers Inc. [10.1109/MELECON56669.2024.10608570].

Neural Network Approaches for State of Charge Prediction of Rechargeable Lithium Polymer Batteries

Forconi F.;Palermo M.;Riganti fulginei f.;Sabino L.
2024-01-01

Abstract

Due to long cycle life and high power and energy density, Lithium-ion batteries have become the main energy storage device in several applications. Since a direct measurement of the battery State of Charge (SoC) is still difficult in many situations, a SoC accurate estimation is pivotal for controlling the battery status and ensuring its maintenance with the aim of reducing operational costs and preventing malfunctions and damages. In view of this, the aim of this paper was to estimate the SoC of rechargeable lithium polymer batteries during charge and discharge cycles by using two different Artificial Neural Networks (ANNs), i.e. Feedforward Neural Network (FNN) and Recurrent Neural Network (RNN). The ANNs were trained and tested with experimental charge and discharge cycles acquired with a home-made automated system. Then, the goodness of prediction of the two approaches was compared in terms of mean-squared error and relative error. Results demonstrated that for the charge cycle the SoC prediction was better with the Recurrent Neural Network approach, with a relative error lower than 5 %. On the other hand, for the discharge cycle the SoC prediction was quite similar between the FNN and RNN.
2024
Apa, L., Del Prete, Z., Forconi, F., Palermo, M., Riganti fulginei, F., Rizzuto, E., et al. (2024). Neural Network Approaches for State of Charge Prediction of Rechargeable Lithium Polymer Batteries. In 2024 IEEE 22nd Mediterranean Electrotechnical Conference, MELECON 2024 (pp.236-241). Institute of Electrical and Electronics Engineers Inc. [10.1109/MELECON56669.2024.10608570].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11590/490612
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact