[Objective] Aiming at the wind power grid-connected power fluctuation problem, this paper proposes a wind turbine rotor kinetic power smoothing strategy based on extended Kalman filter (EKF) and linear active disturbance rejection control (LADRC), in order to solve the problems of phase lag and insufficient disturbance rejection capabilities of the traditional methods. [Methods] Firstly, the aerodynamic model and mathematical model of the permanent magnet direct drive wind power generation system were established. Secondly, the LADRC was designed for the speed loop control, and the rotor kinetic energy algorithm based on EKF was proposed to dynamically update the noise statistical characteristics and optimize the power command estimation. Finally, the simulation model was constructed, and the proposed EKF-LADRC strategy was compared and analyzed with the traditional proportional integral (PI) control and active disturbance rejection control (ADRC) under the turbulence conditions of the high wind speed region and the rated wind speed region. [Results] The simulation results showed that, under turbulent condition in the high wind speed region, EKF-LADRC reduced the power standard deviation by 87.5% and 69.5%, respectively, compared to PI control and ADRC, and suppressed the speed fluctuation to 0.02 r/min. And the power output smoothing was significantly improved under turbulent condition in the rated wind speed region. Simulation results verified the effectiveness of the proposed strategy in suppressing power fluctuations under different wind conditions. [Conclusion] The synergistic strategy of EKF-based rotor kinetic energy algorithm and LADRC proposed in this paper can significantly reduce the power fluctuation and enhance the system robustness. The adaptive noise estimation capability of EKF successfully eliminates the phase lag in traditional filter, and LADRC disturbance compensation mechanism ensures rapid response to turbulent wind conditions. The strategy not only outperforms the traditional methods in terms of dynamic performance, but also maintains operational stability under extreme operating conditions, and its computationally efficient and engineering realizability are of great value for applications in high percentage renewable energy grids that require stringent power quality standards.