Patterns

•A new approach to generate high-quality large-volume battery aging datasets•Our data-generating method saves up to 90% of the battery aging experimental time•The generated datasets exhibit an ultra-low error bounded of 1% only•The first attempt to regenerate data from industrial calibrations like EV maintenance Battery health assessment is crucial for future societal sustainability but it requires extensive and considerable data-sets with a diverse range of usage scenarios. Access to sufficient battery aging data is challenging due to both the complicated battery operations and also the significantly laborious and time-consuming battery aging experiments. We here highlight the potential of combining industrial data with accelerated aging tests for rapid aging data acquisition. We present a machine learning approach to generate high-fidelity large-volume datasets with only 1% error, while the required experimental time can be saved by up to 90%. Our new solution can significantly improve the data-shortage issues for battery aging assessment. The regenerated dataset with representative battery aging dynamic information can benefit many data-based applications such as battery health estimation, lifespan prediction, second-life utilization, and fault diagnostics. Batteries are crucial for building a clean and sustainable society, and their performance is highly affected by aging status. Reliable battery health assessment, however, is currently restrained by limited access to sufficient aging data, resulting from not only complicated battery operations but also long aging experimental time (several months or even years). Refining industrial datasets (e.g., the field data from electric vehicle batteries) unlocks opportunities to acquire large-volume aging data with low experimental efforts. We introduce the potential of combining industrial data with accelerated aging tests to recover high-quality battery aging datasets, through a migration-based machine learning. A comprehensive dataset containing 8,947 aging cycles ...

Recovering large-scale battery aging dataset with machine learning