A brief review is given of the development history of the Ensemble Kalman Filter (EnKF) assimilation method and its essential theory. A numerical test was conducted using the one-dimensional nonlinear wave equation. Results indicate that the method has advantages of being simple to realize and being able to avoid the computation inaccuracy occurred in the prediction process using the EKF covariance variation equation and the problem of massive data storage resulted from the covariance matrix; moreover, the EnKF can control effectively the augment of the estimated error variance for model variables and improve the prediction precision.