Online Capacity Estimation for Automotive Lithium-Ion Cells Incorporating Temperature-Variation and Cell-Aging 2017-01-1191
This work provides a new method for estimating the capacity of an automotive Lithium-Ion cell under real application conditions present in Hybrid and Electrical vehicles. Reliable online capacity estimation is needed for accurate prediction of the remaining electrical driving range. This is a crucial criterion for customer acceptance of Electrical vehicles. Dynamic excitations of real driving cycles, temperature variation as well as the variation of electrical battery behavior with capacity and resistance degradation are challenges that need to be overcome.
For this paper, a long-term aging study on 120 automotive Lithium-Ion cells is evaluated with respect to the correlation between electrical cell behavior, temperature and the cell capacity over the complete cell lifetime. The results are used for a dynamic state-space model which provides the current-voltage relationship valid for all aging states of the battery. The dependence of cell capacity on ambient temperature variation and charge/discharge rate is also analyzed, as well as open-circuit-voltage characteristics with fading capacity.
Using this model, a novel online estimation approach for the cell capacity based on a Quasi-Monte-Carlo-method with multiple Kalman-Filters in parallel is presented. It allows accurate determination of the actual capacity under arbitrary dynamic excitation at different ambient temperatures and unknown aging condition. The new estimation approach offers significant improvements over existing methods due to the inherent separation between long-term estimation of capacity fade and short-term effects of resistance and State-of-Charge.
The method has been proven with real-life driving cycles and offers enhanced capacity and State-of-Charge estimation under full application conditions.