It is known that the model uncertainties will affect the accuracy of state estimation results

which even bring serious consequences. A novel method is proposed for dynamic state estimation of generators with model uncertainties. Firstly

the dynamic state estimation model is developed based on the dynamic equations of generators

and then the dynamic state estimation method based on the adaptive H∞ extended Kalman filter(AHEKF)is proposed. In this method

based on the H∞ filtering theory

the constraint criterion of model uncertainty is established

and the adaptive technique is used to dynamically adjust the covariance matrix and process noise covariance matrix

therefore it has high estimation precision and robustness. Finally

the performance of the proposed method is compared with H∞ extended Kalman filter(HEKF)algorithm and extended Kalman filter(EKF)algorithm both in WSCC 3-machine 9-bus system and an actual large regional power grid system. Simulation results show that

compared with HEKF and EKF algorithm

the AHEKF algorithm performs better in estimation precision and robustness to model uncertainties.