Fig. 4: Analysis of autoregressive prediction results of battery operating features in a charging cycle.
From: Early prediction of lithium-ion battery degradation with a generative pre-trained transformer
a represents the absolute error of battery operating features over 200 timesteps for Long Short-Term Memory network (LSTM), Transformer and Generative Pre-trained Transformer (GPT-Small) model, with the color density indicating error levels. b offers a detailed view of the predictions of baselines and GPT-Small model from different phases (Early, Mid, Late) in a charging cycle. Source data are provided as a Source Data file.
