Abstract: This paper proposes a model predictive control strategy of wind-storage combined operation to solve the problem of gridconnected power fluctuation due to randomness and seasonal imbalance of distributed wind power generation. A wind-storage integrated operation model was established with the objectives of minimizing grid-connected power fluctuations, optimizing storage output, and minimizing energy losses, and Matlab quadratic programming was used to solve the grid-connected power of the wind power and the output power of the energy storage. In addition, a model predictive control algorith is used to optimize the charging and discharging power of battery storage to achieve real-time rolling optimization of the wind-storage grid-connected power, which maintains the storage SOC within a reasonable range and reduces the number of battery overruns. Finally, the example is selected to optimize the operation of a distributed wind farm in Inner Mongolia with different time scales of 1 and 10 min in the annual measured output power recording data.The simulation results show that the proposed method can optimize energy storage charging and discharging in real-time to reduce the grid-connected power fluctuations further and improve the economic operation of the combined wind and storage system.