Estimate state of charge of power lithium-ion batteries used on fuel cell hybrid vehicle with method based on extended Kalman filtering

Abstract

One of the most important tasks of the power lithium-ion battery management systems used on the fuel cell hybrid vehicles (FCHVs) is to precisely estimate the current state of charge (SOC) of the batteries online. A simple physical model and a complex physical model of the battery pack used on FCHVs are proposed, which describe the dynamic behavior of the battery pack by different equivalent circuits respec- lively, and the relationships among different physic variables and state variables of the two models are expressed by the discrete-time state-space functions. Then, based on the Extended Kalman Filter and these two models, the state of charge for the battery pack is estimated by using data such as current and voltage which are sampled when the vehicle is running, and the differences between the two results estimated are also analyzed by using two different models. Our results show that, model based Kalman filter SOC estimation for batteries used on FCHVs is effective. It is also shown that estimating accuracy depends highly on the selected model, and a better model can provide a higher accuracy.