Data and code for temperature prediction of key components in high-speed rail electrical cabinets using LSTM

This record contains the experimental temperature data and author-generated Python code associated with the article “Research on temperature prediction model for key components in high-speed rail electrical cabinets based on LSTM” (manuscript number: PONE-D-26-17701R1). The archive is deposited in its original project structure and includes the following contents: - ‘datas.xls’ contains the temperature data collected during the real-vehicle experiment, including time and temperature values.- The first Python program implements the initial LSTM temperature prediction model. Its corresponding model outputs and evaluation results are stored in the ‘Results’ folder.- The second Python program implements the parameter optimization experiments. The corresponding files and results are stored in the ‘experiment1’ and ‘experiment2’ folders.- Files in ‘.h5’ format are saved LSTM model files generated during model training and parameter experiments.- Files in CSV format contain model evaluation and experimental results.- Files in PNG format contain the corresponding result figures and visualizations. The programs were developed using Python and require relevant packages including TensorFlow/Keras, Pandas, NumPy, scikit-learn, Matplotlib, and Seaborn. These files are provided to support access to the experimental temperature data, model training and evaluation procedures, parameter optimization experiments, and the results reported in the associated article.