In order to enhance the effect and performance of the active distribution network(AND)control, based on ADN Feeder Control Error theory, a distributed model predictive control method for ADN area coordinate control is proposed. Firstly, an autonomous area control model containing intermittent energy resources, load and storage units within the area are established to consider the short-term trend of power output and other forecasting information. Then distributed control method is proposed to solve the coupling issues of optimization targets among different areas, and realize the coordinate control among different areas. Online rolling optimization and real-time measurement feedback are used to achieve the overall optimal control with future forecasts, therefore the global optimization of the whole feeder is assured. Calculation simulation and comparison of the proposed control method with the traditional PI control verify the effectiveness of the proposed method.