Uncertainties on both the source and de-mand side pose challenges for the day-to-day management of distributed power systems, such as microgrids (MGs). To harness the potential for flexible scheduling within MGs and to explore the economic and environmental advantages of cooperative interactions between multiple MGs, this paper develops a two-stage power-sharing model for day-ahead and intraday operations across multi-MGs. The intraday stage, which takes into account the prediction errors caused by uncertainty, refines the day-ahead results by factoring in real-time fluctuations in renewable energy sources and loads, changes in equip-ment status, and the scheduling of various demand re-sponse resources. At the end of the intraday optimal scheduling process, the profits from MGs are allocated according to a generalized Nash equilibrium. A case study involving three MGs within a distribution network, ex-amining both island and cooperative operation strategies, demonstrates the practicality and effectiveness of the proposed approach.