Battery aging datasets for "Increasing generalization capability of battery health estimation using continual learning approach"

These datasets contain different battery types (pouch and prismatic) that have been aged at various temperatures and loading profiles. The partial Q curves during charging and capacities are provided.

Please cite our paper "Increasing generalization capability of battery health estimation using continual learning approach" if use these datasets.

Please note the charge capacity in the datasets is the amount for the ampere-hour throughput in charging mode, which means the capacity in Dataset1 is accounted for with also the charging pulses in dynamic discharging. The charge capacity and discharge capacity are recorded by Neware battery tester. Therefore, it represents the total charging amount during each full charge-discharge cycle. For simplicity and easy use, the SOH curves are normalized by the charge capacity for verification in this work.

Please refer to "The explainable uncertainty in degradation process: a discovery from non-accelerated batteries degradation experiment" for the detailed information and the raw data used as Dataset 5 in the verification. The authors also sincerely thank Dr. Dongzhen Lyu for sharing and explaining the data.