Abstract: In order to coordinate and optimize the economic benefits between the entire AC/DC hybrid microgrid and each region, a partition bi-level optimized operation method of AC/DC hybrid microgrid based on scene analysis is proposed. First, the expected rescheduling cost based on the scenario analysis method is introduced to characterize the random fluctuation of wind and solar charge power. Then, a partition bi-level optimized operation model of the AC-DC hybrid microgrid partition is established, in which the entire microgrid is taken as an upper-level leader, the distributed generation(DG) equivalent total power generation in the AC and DC areas and the purchase-sale of electricity from large grids are taken as upper-level decision variables, the AC area and DC area are taken as two subordinates, the DG output of each region is taken as a lower-level decision variable, and the minimum comprehensive operating cost of each layer is taken as a targeted function. In addition, the improved chaotic black hole particle swarm optimization algorithm is combined with the Gurobi optimization solver to solve the partitioned two-level optimization operation model. Finally, a simulation example verifies that this method takes into account the optimization of the overall AC-DC hybrid microgrid and the economic benefits of each region, and it has a stronger adaptability to the random fluctuations of wind and light loads.