Optimization of Day Ahead Distributed Intelligent Decision-Making for a Multi-Microgrid System

The aim of this paper is to schedule a day ahead operation of multiple grid connected microgrids using two stage optimization scheme. In the first stage, each microgrid forecasts its own load and determines shortage or surplus of power at each hour. In the second stage, power exchanges between microgrids and the utility grid is considered and the microgrids sell their excess energy or buy their shortage energy from the utility grid or other microgrids. Multi-layer feedforward perceptron method is used for forecasting electric loads. Also, stochastic programming is considered in order to account for the uncertainty of power produced by renewable generation units. It was found that by using this energy management method, the total costs of the system were decreased by 2.6%.