Discharging Stage Prediction of Lithium Ion Battery Using CNN - DenseNet Architecture

Lithium-ion batteries are one of the most widely used types of batteries in various applications, including electronic devices, electric vehicles, and energy storage systems. However, battery discharging caused by parasitic reactions between electrodes and electrolytes is one of the most common problems. New methods such as Convolutional Neural Networks (CNN) can help predict discharging modes and improve battery reliability and safety. CNN is used to predict discharging in lithium-ion batteries based on data collected during battery monitoring experiments at different load levels. The data is processed and modeled to support the predictive model. The methodological process involves data collection, data preprocessing, feature selection, and artificial intelligence model training. Model performance is tested using Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) as metrics. The developed CNN model is observed to follow data trends. For the CNN model, an RMSE of 0.0255, MAE of 0.0182, and MAPE of 0.216% were obtained during the training process on 80% of the data for voltage prediction. The model has been tested to be applicable in real-world scenarios for detecting battery discharging.