[Objective] To address high operational costs and non-smooth power exchange with the grid in existing campus microgrids that rely solely on power sources or air-conditioning loads to mitigate fluctuations，we propose a multi-timescale coordinated scheduling strategy for micro-turbines （MT） and air-conditioning building clusters in campus microgrids considering uncertainties in grid curtailment. [Methods] First，we established a virtual energy storage model for air-conditioned buildings by quantitatively analyzing the flexible energy characteristics of variable-frequency air conditioners. Combined with the relationship between power generation and energy consumption for MTs，we propose a short-timescale efficiency-coordinated control method for MTs and variable-frequency air conditioners to adaptively mitigate power fluctuations. Subsequently，a long-timescale coordinated operation model between MT systems and air-conditioned buildings was constructed based on supply and demand relationships relating to the microgrid. Furthermore，robust optimization methods were implemented with stochastic scenario exploration to characterize uncertainties in grid curtailment，renewable generation，and load demand. Overall，we aimed to design a flexible multi-timescale scheduling strategy compatible with grid-connected and islanded modes. Finally，we linearized the model into a mixed-integer linear programming （MILP） problem. [Results] Simulation results demonstrate that the system strictly followed the efficiency-coordinated control method to dynamically coordinate power allocation between the MT and virtual energy storage when the power provided by the microgrid fluctuated. The proposed strategy reduced the total system costs by 10.1% in the grid-connected mode and 5.0% in islanded mode. It mitigated risks of systemic collapse effectively while enhancing the reliability and economic performance of the system. [Conclusion] The proposed strategy takes advantage of the coordinated scheduling potential of MT and variable-frequency air-conditioning clusters to adapt flexibly to grid-connected and islanded scenarios. It significantly improves the economics of microgrid systems，and provides an efficient scheduling solution for high-penetration renewable energy integration.