In this paper, an ML-enhanced testing procedure for li-ion batteries is proposed, which increases the efficiency of cycle life testing by stopping the cycling of a subset of batteries and predicting their end of lifes instead. Based on aging features and the Kennard-Stone algorithm, a suitable subset of batteries is selected, whose future aging trajectory can be predicted with high accuracy. The cost saving potential of the novel testing procedure is shown on the basis of three different cycle life studies.
Increasing the efficiency of li-ion battery cycle life testing with a partial-machine learning based end of life prediction / Thomas Kröger, Alexander Bös, Sara Luciani, Sven Maisel, Markus Schreiber, Markus Lienkamp / Journal of Energy Storage, 2023, 73 (Part A), 108842 – DOI