The coordinated operation of “source-grid-load” considering the high dual uncertainties of the high proportion of distributed photovoltaic output and the massive flexible loads is a difficult problem that needs to be solved urgently in the development of new power system. For this reason, a deterministic optimal model for the economic autonomous operation of the transformer area with the goal of minimizing operating costs considering the flexible adjustability and demand response characteristics of active loads is established; Based on this, a robust polyhedral uncertainty set that characterizes source-side and load-side fluctuations is constructed, and an uncertainty optimization model for economic autonomous operation of transformer areas is established. Then, the autonomous uncertainty optimization model is converted into a robust optimization model of decoupling iterative solution to realize the distributed iterative solution by using ADMM algorithm. It can be concluded from simulation and comparison experiments that the solve efficiency of the distributed optimization method that takes into account the dual uncertainty of sources and loads for the economic operation of the transformer area is better than the centralized algorithm and can promote the distributed photovoltaic consumption while reducing the users’ electricity costs.