State of Charge Estimation for Lion-Lithium Batteries Using Extended Kalman Theorem

For the hybrid and electric vehicles, a battery management system is required to supervise and optimize the electrical energy management. SOC and SOH are key parameters to supervise the good use and to contribute to enhance the lifetime of the battery. This paper presents the research for an intelligent state of charge estimation. The state of charge is a very important factor to characterize the state of the storage elements. This work consists in implementing a real time observer in order to supervise key parameters of the battery. An extended Kalman filter based on a lumped battery model is used to achieve this goal. A lumped model has been defined to consider the electrochemical phenomena, and has been linked with a thermal model to take into account the influence of the temperature. A strategy to set up the Kalman filter has been proposed to get a fast and accurate convergence of the state of charge observation despite bad initializations. Extensions Kalman have been implemented to take into account the influence of the aging on the battery's dynamic model. For the hybrid and electric vehicles, a battery management system is required to supervise and optimize the electrical energy management. All the simplified in order to be implementable on an embedded digital signal models have been processor. The real time state of charge observer has been validated on a lithium ion cell for different operating conditions.