Data-Driven Battery Lifetime Model Calibration and Analysis for an Electric Vehicle Battery’s Durability Performance 2024-01-2281

Due to the expense and time commitment associated with extensive product testing, vehicle manufacturers are developing new simulation techniques to verify vehicle component performance with less testing and more confidence in the final product. Battery lifetime is of particular difficulty to predict, since each battery is different and there are many different control scenarios that could be implemented based on the specific requirements of each battery type. In order to solve this problem for a 12V auxiliary lead-acid battery, a battery durability analysis model has been previously adapted from lithium-ion applications, which is capable of verifying the impact of lead-acid battery durability in a short period of time. In this study, calibration tools for this model were developed and are presented here, and durability analysis and verification are performed for the application of new electric vehicles. New control strategies, designed specifically for the auxiliary batteries in electric vehicles, were evaluated and the durability guide standard (current specification within 6%) was met. The current simulation tools are set up based on battery unit tests, and coefficients are adjusted to account for the impact of measurement procedures on measurements in a full, end-to-end, battery lifetime simulation process that takes an engineer from raw data all the way through final lifetime prediction. The tools are built to simulate expected battery life under different environmental and charge control scenarios. EV drive cycles were run to determine whether or not flooded (FLD) or absorbed glass mat (AGM) batteries were more appropriate for the application studied here, and whether each battery type would be able to meet warranty requirements. Results show that an AGM battery would not fail in a 5 year usage interval, and that a FLD battery would not meet the minimum 3 year requirement for the control scenarios used.