Abstract

The paper describes the application of observer-based state-estimation techniques for the real-time prediction of state-of-charge (SoC) and state-of-health (SoH) of lead-acid cells. Specifically, an approach based on the well-known Kalman filter, is employed, to estimate SoC, and the subsequent use of the EKF to accommodate model non-linearities to predict battery SoH. The underlying dynamic behaviour of each cell is based on a generic Randles' equivalent circuit comprising of two-capacitors (bulk and surface) and three resistors, (terminal, transfer and self-discharging). The presented techniques are shown to correct for offset, drift and long-term state divergence-an unfortunate feature of employing stand-alone models and more traditional coulomb-counting techniques. Measurements using real-time road data are used to compare the performance of conventional integration-based methods for estimating SoC, with those predicted from the presented state estimation schemes. Results show that the proposed methodologies are superior with SoC being estimated to be within 1% of measured. Moreover, by accounting for the nonlinearities present within the dynamic cell model, the application of an EKF is shown to provide verifiable indications of SoH of the cell pack.