Modeling and planning of energy production in renewable energy stations with artificial neural networks

Wind power is the most common, widely applicable, feasible and productive renewable energy source. However the control of the wind power stations is mostly under control of the nature. In other words, wind speed and power production can not be directly controlled by human beings. One of the major problems in the process of electricity generation from renewable energies, which includes the timing from planning to production, is the modeling of meteorological activities and corresponding production level. Complicated meteorological and mathematical models can not provide correct and flexible results. If the production level of the renewable energy plant can not be defined correctly according to the meteorological data then hardware capacities, investment and production costs of the plant may calculated erroneously. In the national level, fluctuations and interrupted electricity power entry to the national grid result in technical problems, decreases in capacity and feasibility problems. This thesis proposes a new approach for solving the above mentioned problems by the use of artificial neural networks. A model for forecasting the wind, hydro and solar radiation capacity for short, mid and long term has been developed. As a result of this, energy production planning and control in the renewable energy plants can be achieved in daily, monthly and yearly basis with great senility and straightness. This new model and approach was also tested in different regions of Turkey with real data obtained from meteorological data centers. Keywords: Renewable Energy, wind power, hydroelectric power, fuzzy logic, artificial neural networks, energy planning.