Related Experiment Video
Updated: Oct 5, 2026

The Effect of Charging and Discharging Lithium Iron Phosphate-graphite Cells at Different Temperatures on Degradation
Published on: July 18, 2018
Industrial LiFePO₄ battery management system dataset for early degradation analysis and remaining useful life
Mst Asfia Binte Ulfat1, Afsana Begum1, Md Abdulla Al Mamun2
1Department of Software Engineering, Daffodil International University, Dhaka, 1216, Bangladesh.
Abstract:
This data article describes an experiment-informed R&D dataset associated with an industrial LiFePO₄ battery pack with a nominal voltage of 25.6 V and a rated capacity of 40 Ah. The dataset was prepared with industrial support from GMP Lithium Ltd., Bangladesh, and represents battery operational behaviour and capacity degradation within practical industrial operating ranges. Capacity(Ah) follows a deterministic degradation trajectory calibrated to the manufacturer's ageing-test end points, whereas the pack-level operating variables and the eight cell voltages are structured (simulated) cycle-level values generated within the manufacturer's specification limits; the complete generating equations are reported. The dataset consists of 3000 sequential charge-discharge cycle records in CSV format. Each record contains pack identification, cycle number, battery capacity, State of Health, Remaining Useful Life, pack voltage, pack current, a normalized lifecycle progress indicator labelled DoD(%), operating temperature, and individual voltage measurements from eight series-connected cells. The dataset contains 17 variables, including indicators of battery degradation, pack-level operating conditions, and cell-level voltage behaviour. Supporting documentation, including a codebook, README file, and data publication permission letter, is provided with the dataset in the Mendeley Data repository. The dataset can be reused for battery degradation analysis, State of Health estimation, Remaining Useful Life prediction, assessment of cell-voltage imbalance, feature analysis, and development and evaluation of machine learning and deep learning models. The combination of capacity, operational, and cell-level voltage variables enables the study of battery ageing at both pack and cell level. It is not a raw continuous BMS log and was not sampled directly from a physical test bench; it should not be interpreted as such.