Sign in Subscribe
Artificial intelligence

Breakthrough in Energy Storage: Researchers Develop Innovative Model for Battery Pack Capacity Prediction

Strategic Insights

— 1 min read

Researchers at Nantong University in China have made a significant contribution to the field of energy storage by developing an innovative model for predicting battery pack capacity degradation paths. The model, which integrates Convolutional Neural Network (CNN) and Fourier Neural Network (FNN) architectures, has demonstrated superior performance in predicting battery pack capacity degradation compared to other models.

Key Takeaways:

  • The research, titled "The Integration of the Convolutional Neural Network and Fourier Neural Network Methods for the Battery Pack Capacity Prediction," was published in the Journal of Energy Storage in 2025.
  • The model integrates traditional CNN architectures with FNN to capture both time-domain and frequency-domain features of input data.
  • The channel attention mechanism is employed to derive weight information from input features, and one branch of the input undergoes mapping to a high-dimensional space through fully connected layers to extract frequency-domain information.
  • The other branch utilizes a CNN network to grasp the temporal characteristics between features, and time and frequency domain information are extracted through one-dimensional convolution.
  • The fused prediction is achieved through two fully connected layers, and the efficacy of the proposed method is validated using charging data from 20 Electric Vehicles (EVs).
  • The proposed model surpasses other models, including its two submodels, and exhibits consistent high performance across varying prediction windows.
  • The research has been peer-reviewed and is suitable for enhancing the prediction of battery pack capacity degradation paths.

Statistics:

  • 20 Electric Vehicles (EVs) were used to validate the efficacy of the proposed model.
  • The model demonstrated superior performance compared to other models, including its two submodels.
  • The proposed model exhibits consistent high performance across varying prediction windows.
  • The research was published in the Journal of Energy Storage in 2025, issue 125.

Sources:

  • NewsRx LLC, "Researchers at Nantong University Target Energy Storage (The Integration of the Convolutional Neural Network and Fourier Neural Network Methods for the Battery Pack Capacity Prediction)," Energy Weekly News, August 8, 2025, p 810.
  • The Integration of the Convolutional Neural Network and Fourier Neural Network Methods for the Battery Pack Capacity Prediction, Journal of Energy Storage, 2025;125.
  • Elsevier, Radarweg 29, 1043 Nx Amsterdam, Netherlands.
  • Xingxing Wang, Nantong University, School of Mechanical Engineering, 9 Seyuan Rd, Nantong 226019, People's Republic of China.

Read more