The conventional automatic cut-out strategy for offshore wind farms during typhoons leads to large-scale shutdowns and triggers significant wind power ramping events. To address this issue

this paper proposes a spatiotemporal optimization scheduling strategy for the orderly avoidance of these ramping events. First

considering that the accuracy of typhoon prediction is difficult to apply to the assessment of wind conditions

a downscaled typhoon simulation scheme is designed using WRF to meet the needs of avoidance scheduling. Then

in response to the differences in operational features of wind turbines during the evolution of typhoon paths

complex risk avoidance strategies are formulated. A directed wake graph of offshore wind farms is constructed to achieve dynamic spectral clustering

avoiding the disorder of wind turbines automatically cutting out. Finally

in response to inaccurate description of wind power fluctuations under climbing events

the ramping feature is introduced to improve the uncertainty set based on Wasserstein distance

and a robust optimization model for risk avoidance distribution is established. The rationality and effectiveness of the proposed model is verified through numerical simulations.