The use of renewable energy is now a key focus of research in order to solve the increasingly prominent environmental and energy problems. Microgrids are widely used because of their high utilisation of renewable energy. However, renewable energy is intermittent, fluctuating and difficult to predict accurately, and the load demand is also uncertain, therefore, this study focuses on the ADMM algorithm to accomplish the optimal scheduling among distributed networks. While for individual microgrids BP neural network is used for prediction to derive the optimal solution for microgrids under disturbances.