Wind power has obvious volatility and anti-peak shaving characteristics. To reduce the impact of large-scale wind power grid connection on power system peak shaving, this paper proposes a multi-time scale scheduling scheme based on wind power peak shaving scenarios. This improves the implementation efficiency of a power grid scheduling scheme in extreme wind power scenarios. First, this paper constructs a wind power load data-driven model, and then a multi-time scale evaluation index of wind power peak shaving power is proposed. From this index, the wind power of a year is decomposed in the time domain and typical wind power peak shaving scenarios are generated for analysis. Second, in order to ensure the power balance of the power grid in the peak shaving scenario, a wind-thermal-energy storage coordinated peak shaving model is built for the historical scenario, and the economic optimal dispatching plan for typical scenarios is formulated. Then a multi-time scale scheduling scheme is proposed to deal with the uncertainty of wind power through rolling revision of the scheduling plan. Finally, the analysis of an example shows that the proposed method can effectively improve the calculation speed of the dispatching scheme and implement the dispatching scheme in time to ensure economic and stable operation of the power grid.