To fulfill the optimal allocation of an active distribution network to act as an effective management platform for distributed generation, an optimal allocation model based on robust optimization theory is established. First, by taking the comprehensive allocation cost as the objective function, and considering power balance constraints, controllable micro-source operation constraints, rotating reserve constraints and other necessary constraints, an active distribution network optimal allocation model is established. In view of the uncertainties in the model, a robust optimization framework is adopted to eliminate the uncertainties based on the distributed photovoltaic and wind power interval models. Linear duality theory and the Lagrange derivative method are used to transform the uncertainties into a deterministic optimal allocation model. The simulation example shows that the active distribution network has higher operational economy and reliability than the traditional distribution network. The model established in this paper can take into account the impact of uncertainties of wind and solar output on the configuration results, and so it can be applied to the formulation of the optimal configuration scheme of an active distribution network. This work is supported by National Key Research and Development Program of China (No. 2018YFB2100100).