Abstract: The randomness and fluctuation of distributed generators（DGs） output and the uncertainty of measurement error distribution bring challenges to the accurate situation awareness in the AC/DC distribution network. Aiming at obtaining the entire interval of system states under the influence of multiple uncertainties, this paper proposes a three-phase interval state estimation（SE） model for AC/DC active distribution networks based on the description of DG output and measurement values using the interval forms. In order to obtain an accurate initial state interval, the state interval at the current moment is predicted by using the historical state, and the interval correction is performed based on the nonlinear optimization model to ensure the reasonableness of the predicted interval. Then, the nonlinear interval SE model is transformed into a linear optimization model by using a nonlinear equation deflation method based on the median theorem to ensure the completeness of the estimated intervals. Finally, the effectiveness of the proposed method is verified by simulation.