Experimental based Aging Model for Automotive Li-Ion Batteries

The work provides a study on the development of algorithms which estimate the degradation rate of battery cells undergoing an extensive aging campaign lasting approximately 2 years. As the cell is progressively used, different degradation mechanisms might take place. Power and Energy fade, together with loss of reliability and safety, are the main consequences of cell degradation. In this work an aging model has been devised for the estimation of resistance increase and capacity loss as function of the cell-working condition. The modelling has been performed in MATLAB by data fitting the results obtained from the experiments using optimization algorithms such as Levenberg-Marquardt. The proposed aging model is combined with thermal and electrical models, which estimate the cell working condition of temperature and state-of-charge starting from the battery current profile measured during operation. A real driving cycle validated the goodness of the results by closely matching the prediction of the resistance increase and capacity loss with a coefficient of determination (R2) value of 0.9476 and 0.9262, respectively. The proposed model will contribute to the improvements of vehicle operation and the assessment of energy management strategies developed on a virtual-only basis, thus lowering costs and improving prediction accuracy.