State of Charge Estimation for Series-Connected Lithium Battery Pack Using Extended Kalman Filter

The battery pack is one of the main components in electric vehicles which is usually composed of many cells connected in series. Battery state of charge estimation is one of the most important functions of the battery management system in electric vehicles. Due to the different manufacturing and operational conditions, all of the cells in a battery pack do not have the same states of charge and therefore, the cell and pack states of charge are not the same. This paper presents a method for battery pack state of charge estimation which benefits rather low computational cost as well as the high precision. First, the coulomb counting method and the open circuit voltage graph, which is obtained from the experimental results, are used simultaneously to estimate the pack average state of charge. Then, the extended Kalman filter method is used to estimate the difference between the pack average state of charge with those of the cells. The proposed method has been evaluated and verified using an experimental test bench for three series-connected lithium cells. Experimental test results indicate good performance of the proposed method in estimating the lithium battery pack state of charge.