Cao, GuozhongSun, ShijingNeelgund Ramesh, Hemanth2024-10-162024-10-162024NeelgundRamesh_washington_0250O_27552.pdfhttps://hdl.handle.net/1773/52558Thesis (Master's)--University of Washington, 2024Lithium-ion batteries (LIBs) undergo irreversible and complex aging processes, making it essential to evaluate how different parameters influence their cycle life. While previous studies have examined ageing behavior by changing individual cycling parameters, the cumulative effects of multiple parameters remain largely unexplored. Therefore, in this study, we employ explainable machine learning techniques on 28 LIBs from publicly available data to gain comprehensive understanding contribution and interaction of various parameters towards degradation as the battery ages. Our results suggest that in the early stages of battery life (State of Health (SOH) between 1.00 and 0.90), the degradation is sensitive to change in temperature, which is likely due to the abundance of available electrolyte. As the battery ages (SOH between 0.875 and 0.800), charging current emerges as the primary degradation factor. This shift is attributed to the development of the solid electrolyte interphase (SEI) and the subsequent decline in electrolyte quality, which leads to uneven charge density distribution and localized thermal effects, such as Joule heating. Additionally, the charging current and temperature show large interaction effects causing degradation. These findings are consistent with existing literature regarding changes in degradation mechanisms over the aging period.. By offering a qualitative analysis of the differential impact of various parameters on cell aging, our study provides a novel understanding of battery degradation mechanisms. This work lays a solid foundation for future research aimed at extending the cycle life and improving the safety of LIBs using data-driven methods.application/pdfen-USCC BYbatteriescyclingdegradationmachine learningmodelingMaterials ScienceMaterials science and engineeringThesis