Smart management of micro-grid systems which containing renewable energy resources

In this thesis study, intelligent management of a microgrid with distributed storage unit consisting of solar and wind energy plants, energy from waste (EfW) plant and pumped storage hydroelectric power plant (PHES) is aimed. In this direction, energy balancing and storage processes are defined by the conditional flow algorithm and autonomous grid management system is developed. Successful management patterns, which are generated autonomous grid management system was used in the training of extreme learning machines (ELM), and intelligent management of the microgrid was realized by ELM. Proposed methods were tested for different production and consumption scenarios by simulation models developed in Matlab and PowerWorld environments and the results obtained were discussed. In the simulation studies, models of production profiles of renewable resources consisting of solar and wind power plants and EfW, behavioral model of distributed storage unit and variable demand models including households, hospitals and schools for analysis of different demand conditions have been used. For strong, weak and island mode cases, both PHES and non-PHES conditions have been analyzed, separately. As a result of these analyzes, the energy exchange rates between the microgrid and the distribution grid were calculated for the cases of with or without the PHES, and the data evaluating the microgrid's dependency on the distribution grid was obtained. In addition, the number of renewable energy sources required for the microgrid to operate in the island mode with PHES, where the microgrid does not receive energy under any conditions, but can provide energy to distribution grid, is obtained by an algorithm that uses brute force search method.