Towards real-time (milliseconds) parameter estimation of lithium-ion batteries using reformulated physics-based models

In this paper, transport and kinetic parameters of lithium-ion batteries are estimated using a rigorous porous electrode theory based model. The rigorous model used in this investigation is reformulated using advanced mathematical techniques. Since batteries and other electrochemical devices are used in hybrid environments,which include devices with time constants less than a second (like supercapacitor), we need to develop parameter estimation codes with computation time less than a second or a few milliseconds. In this investigation, the computation time for parameter estimation measures between 100 and 300 ms since a reformulated battery model is devised especially for these purposes. Obtaining the numerical solution for battery model equations is very difficult towards the end of discharge and is usually neglected for parameter estimation purposes. However, in this paper the estimation takes into account the entire discharge data ranging from an initial potential of 4.2 V to a cut-off potential of 2.5 V. It is found from this investigation that the reformulated lithium-ion battery model is efficient and accurate in estimating parameters. (C) 2008 Elsevier B.V. All rights reserved.