A control method for reducing a wind power grid-connected scheduling plan error comprises the steps of firstly analyzing the wind power outlet characteristics in a wide-area time scale and the influence of the wind power outlet characteristics on a power grid scheduling plan, wherein the influence comprises the influence of the wind power output randomness in the wide-area time scale on the power grid scheduling plan, the wind power output fluctuation characteristic in the wide-area time scale on the power grid scheduling plan, and the influence of the wind power intermittent and anti-peak shaving characteristics on the power grid scheduling plan; and then establishing a multi-dimensional scale system of the wind power grid-connected power system scheduling plans on the basis of analyzing the wind power outlet characteristics in the wide-area time scale and the influence of the wind power outlet characteristics on the power grid scheduling plan; establishing mutually coordinated scheduling plans in the multi-dimensional scale system, adding the optimization schemes of other scheduling plans in the scheduling plans in different scales, and utilizing the scheduling plans of different scales to optimize mutually, thereby controlling the reducing the error of the wind power grid-connected actual output and the plan output.