The invention discloses an active power distribution network dynamic reconstruction method based on grid structure similarity. The method comprises the following steps: power distribution system parameters and DG parameters are determined; a static reconstruction model is established by taking maximum DG output as a target in each unit time period, and an optimal DG output value and a grid structure in each unit time period are determined; the grid structure similarities between adjacent time periods are measured, and a merging time period is determined; the DG output is optimized to maximizethe total DG output, and the grid structure which maximizes the total DG output is selected as the optimal grid structure in the merging time period; the number of divided time periods, the grid structure of each time period and a DG output value are updated; and whether to meet the constraint conditions of the switching frequency and the maximum reconstruction frequency is judged, if yes, the process is ended, or otherwise, the process is returned. The invention provides a power distribution network dynamic reconstruction strategy based on the grid structure similarity, which can optimize thegrid structure of the power distribution network, delay the investment of the power grid, improve the renewable energy consumption rate of the system and reduce the power purchase cost to the superior power grid.