Robustness evaluation for state-of-charge and state-of-health estimation considering electrochemical parameter uncertainties

Electrified automotive powertrains benefit from precise knowledge of the battery state-of-charge and state-of-health to aggressively utilize the battery for fuel economy and range improvements while ensuring overall system safety and reliability. Uncertainties associated with the electrochemical parameters that govern the concentration, potential, and reaction rate dynamics within Li-ion cells can lead to state estimation errors and non-optimal battery usage. In this paper, results are presented towards quantifying the effect of parametric uncertainty in an automotive-oriented battery state estimation algorithm. Extensive simulations are conducted via a design of experiments approach to quantify closed-loop robustness and identify electrochemical parameters whose uncertainties create disproportionately large estimation errors. The results indicate that the effects of parametric uncertainty can be minimized by applying closed-loop estimation to the states that exhibit the largest overpotential within the cell.