In recent years, lithium iron phosphate (LiFePO4) power battery is widely used for electric vehicle. However, it is difficult to estimate the state of charge(SOC) of battery because of the characteristics of material itself. In complicated operation environments, SOC estimation plays a significant role in ensuring safety and reliability of battery operations for an electric vehicle. In this paper, both unscented Kalman filter and Particle Filter methods of a LiFePO4 battery for applications in electric vehicles were verified using Thevenin equivalent circuit model. Compared with the extended Kalman filter method, results show that both unscented Kalman filter and particle filter have a better estimation accurancy.