Oka et al., Supplementary Data for "Development of a battery emulator using deep learning model to predict the charge–discharge voltage profile of lithium-ion batteries"

Supplementary data for the paper, Oka et al. "Development of a battery emulator using deep learning model to predict the charge–discharge voltage profile of lithium-ion batteries". (DOI: 10.1038/s41598-024-80371-9 )A zip file contains following data and codes.01_lstm_model_making.pyThis file is a Python script for reading battery charge-discharge data and training an LSTM model. The output is in the PyTorch .pth file format. For a single file, test data is read, and the prediction plot is output. To use this Python script, you need to modify the "CFG (config)" and "Convenient" sections within the script.02_lstm_reuse_pthmodel.pyThis file is a Python script that reads the .pth file output by 01_lstm_model_making.py and predicts the voltage profile from the given current value and the integrated value of current and time (proportional to SOC). To use this Python script, you need to modify the "CFG" and "Convenient" sections within the script.01_dualfoil_simulationThis directory contains Dualfoil simulation program. The "dualfoil.in" file is a representative input file used in this study.02_dualfoil_dataThis directory contains current and voltage profiles created by the Dualfoil simulator (30 training data / 20 test data) and representative LSTM trained model (.pth file).03_experiment_dataThis directory contains current and voltage profiles obtained from the actual battery created (30 training data / 8 test data) and representative LSTM trained model (.pth file).04_hyperparameter_tuning_resultsRMSE of test dataset using prediction functon derived from various hyper parameters.(CSV formatted)AcknowledgementThe battery cells used in the study were manufactured and supplied by LIBTEC.