Abstract: For the development of electric vehicles, power lithium-ion battery is an option of assistant power source for such kind of vehicles. An equivalent physical model of the power lithium-ion battery pack is founded which is then described in a discrete state-space form. The principle of how to use Kalman filtering to estimate the internal states including state of charge and state of health independently is introduced, and based on this, how to estimate state of charge and State of health synchronously by using dual extended Kalman filtering is proposed. A corresponding experiment is designed in which the data such as current and voltage are acquired to test the algorithm, and the testing results show that internal state estimation of battery by the method of dual extended Kalman filtering is effective and results in a relatively hiyh estimation accuracy, thus well reflecting the actual internal state of the battery pack.

Key words: Dual extended Kalman filtering, Electric vehicles, Power lithium-ion battery, State estimation