To alleviate the problem of poor active regulation of wind power itself, this paper proposes a multi-timescale operation strategy for wind storage systems to enhance the active regulation capacity of wind farms. First, in order to reduce the number of energy storage charge and discharge state transitions, the energy storage units are divided into charge and discharge groups to undertake the charging and discharging needs respectively. Then from the state of charge of the energy storage units, a dynamic grouping mechanism is used to update the energy storage units in the charge and discharge groups. Second, a multi-timescale operation model for wind storage systems is constructed based on the model predictive control method. This is used to enhance the active regulation capability of wind farms over different timescales. A power allocation strategy is proposed for energy storage units to refine the management of the orderly action of energy storage units and to optimize the power allocation of energy storage output among the units. Finally, a wind farm in Xinjiang is used as an example. The results show that the proposed strategy can effectively improve the active regulation capacity of the wind storage system and reduce the number of energy storage charge and discharge state transitions.