Jonathan Wirth

Exploring potentials of stimuli for enhanced robustness within a pre-competitive working group, strongly supported by Consortium for Battery Innovation (CBI), a database with test vectors was created in which lead batteries were subjected to daily driving cycles. The scenarios simulated include different vehicle (BEV and µHEV) and usage types (sales agent, short-distance commuter, elderly care, craftsperson, and others), climatic variants and varying generator and alternator setpoints. The batteries tested cover a broad range of field-relevant ageing effects. The aim of the test vectors is to validate algorithms for predicting the performance of the lead battery. For this purpose, the end voltage of a predefined safety-relevant discharge current profile (Safety State-of-Function, SSoF) is continuously predicted based on regular stimuli (ripples, charge/discharge steps). The battery is then occasionally subjected to an actual SSoF profile to validate the prediction. Six different SSoF profiles of different size and duration are used, five of which comprise two steps. In this paper, the stimuli and SSoF profiles of selected test vectors are statistically analyzed to show the extent to which certain properties of the stimuli can contribute to an accurate short-term prediction. As a basis, the correlation of the high-frequency resistance, which represents the ohmic components, to the end voltages of the SSoF profiles is examined and the limits of a corresponding prediction are shown. Based on this, other properties of the stimuli such as voltage behaviour and charge/discharge quantities are considered and the extent to which they can contribute to an improvement in the prediction is examined. The audience will gain valuable insights into challenging operating points for the prediction, what potential lies in the individual stimuli and in which direction further investigations may be worthwhile.