Abstract: The power system frequency response analysis methods based on model-driven or data-driven methods have contradictions between calculation speed, accuracy, and generalization ability. Considering the advantages and disadvantages of various analysis methods in the real-time application,a model-data integration driven analysis method is proposed in this paper. In the modeling process,the method uses System Frequency Response（SFR）model to predict the frequency response dynamic and uses Extreme Learning Machine（ELM） Optimized by Particle Swarm Optimization（PSO）model to correct the error of the predicted results. The model-data integration driven method can greatly improve calculation accuracy while keeping a high calculation speed.Moreover,the method is less dependent on the sample data to improve the generalization ability. The simulation on the WSCC 3-generator 9-bus test system has verified that the method can quickly and accurately calculate the dynamic process of frequency response after a large-scale disturbance,and has good generalization ability. It can further provide an emergency assistance decision for the dispatching and control of the power system to prevent the frequency crash.