Abstract: Parameter identification of lithium battery model is very important for internal state estimation, performance optimization and safe operation of the battery. Aiming at the problem of poor estimation accuracy of traditional Kalman filter(KF) algorithm, a parameter identification method of second-order RC equivalent circuit model of Li-ion battery based on MOPSO-KF algorithm was proposed. The method uses multiple target fitness functions and multi-objective particle swarm optimization(MOPSO) algorithm to quickly determine the optimal KF noise covariance matrix, improve the stability and prediction accuracy of FK algorithm, and then realize the accurate identification of lithium battery model parameters. The results show that MOPSO-KF algorithm is closer to the actual measured value(RMSE less than 0.005 V) in terminal voltage prediction than the traditional method, and has higher data tracking accuracy and fitting accuracy. This method can effectively improve the accuracy of battery model parameter identification, and provide a solid foundation for battery condition monitoring and health management.