The lithium-ion battery plays an important role in saving energy and lowering emissions. Many parameters like temperature have influence on the characteristic of the battery and in an electric vehicle this phenomenon becomes more serious. Estimating these parameters is kind of challenging work and many available approaches have been studied such as Kalman filter and offline AC impedance. In this paper, the application of a boost DC/DC converter to the battery system of high power for online AC impedance identification is proposed. The function of the converter is to inject a current fluctuation signal into the battery at work and the normal output current is drawn by a load. Through analyzing the average state space equations and deriving the small signal model of the converter, the gain function is deduced of the fluctuated current signal against the fluctuated duty cycle which controls the converter. The control algorithm is designed and the system model is verified using Matlab/Simulink with respect to the disturbance current signal generation, the gain function and its variation with frequency range. Then the converter is designed and implemented to conduct the experiment on a power battery of 175Ah and the nominal voltage is 100V. The measured AC impedance of several single cells is presented and it demonstrates the feasibility of applying this DC/DC converter to the practical electric vehicle.