Abstract: As the integration of high proportions of renewable energy sources into the grid increases, the power fluctuations on the supply side increase. During maintenance of the 220 kV side in a substation, the power grid operation faces more uncertainty and risks, which demands higher requirements for robustness in maintenance planning. However, currently there are no mature methods to assist in the development of maintenance plans that can adapt to the complex and changing transmission environment of the 220 kV side in a substation. To address this issue, this paper first builds a deterministic model for substation maintenance planning on the 220 kV side based on the transmission network planning theory and engineering practice experience. The model considers the optimization objective of economic cost and the reliability of the system's operation during maintenance, and takes into account maintenance project constraints and power grid flow safety conditions. Furthermore, by using the probability-driven self-adaptive robust optimization theory, renewable energy output and equipment failures are incorporated as uncertain factors into the aforementioned deterministic model, thus a maintenance planning model for the 220 kV side of a substation that considers the effects of uncertainty factors can be obtained. The model is optimized based on the benders decomposition and linear programming theory to improve the accuracy and speed of the model solution. The effectiveness of the model is verified by analyzing the maintenance planning of an improved IEEE-RTS79 example and a real case in the power grid.