Dynamic equivalence modeling of wind power plant became an important research method for analyzing the grid connection characteristics of large-scale wind power plants. Currently, there had been extensive studies on the characteristics of a single doubly-fed induction generator (DFIG), but there was very little research on the overall characteristics of wind power plant. Due to the huge difference between wind turbines, the multi-machine equivalent model could fully reflect the dynamic characteristics of wind power plant in a better way. To determine the generator parameters of the multi-machine equivalent model, the method of data processing was explained in detail in this paper based on the data measured with the point of common coupling(PCC), and moreover, the improved particle swarm optimization (MPSO) was used to identify the parameters of the equivalent wind turbine on the basis of fully considering the external characteristics. The multi-machine equivalent model of wind power plant (WPP) based on DFIG was established, and the comparison with the refined model of wind power plant showed that the model constructed with the method mentioned in this paper can reflect the dynamic characteristics of wind power plant more accurately, which exactly verified the effectiveness of this method.