ABSTRACT
An intelligent battery sensor is an important component for vehicular energy management, and the precision of online estimation of the state of charge determines sensor’s performance. A simple equivalent circuit was employed to describe the dynamic properties of vehicle lead-acid batteries. A model is reported to estimate the state of charge using the approximately linear relationship of open circuit voltage with this parameter. A time-varying circuit of the model was prepared and an identification model was derived to improve the accuracy. A method that employed a cascaded combination of double improved extended Kalman filters is reported and employed in a demonstration intelligent battery sensor. This approach has higher precision and speed in estimating the state of charge compared to previous protocols.