In order to adapt to the requirement that distributed generations access distribution network with a small quantity of measuring equipment, based on the forecasting of wind power and photovoltaic generation, a weighted least squares state estimation model with constraint condition is proposed by regarding the distributed generations’ power forecasting result and "virtual" reactive power as pseudo measurement. In the model, Jacobian matrix is revised by the reactive-voltage characteristic of asynchronous wind turbine, the equality constraint of photovoltaic power generation system’s zero reactive injection is used to revise the result of state estimation. To ensure the state estimation’s numerical stability, singular value decomposition is used to solve the model. Simulation result of IEEE33 node example indicates that this model integrates distributed generations’ own characteristics with weighted least squares state estimation, which reflects the distributed generation’s output more accurately, and improves the precision of state estimation without addition of measuring installation.