Performing system performance analysis using artificial intelligence methods in photovoltaic systems

With the rapid increase in population and the increasing industrialization in the world, the need for electrical energy is increasing. Nuclear and fossil fuels are generally used to meet this increasing demand. The increase in the use of fossil fuels has led to an increase in CO2 emissions and consequently an increase in air pollution. As a result of air pollution, the precipitation of acid rain, the depletion of the ozone layer and the increase in the effects of global warming disrupt the balance of nature and affect human life negatively. In addition, the use of fossil fuels has increased the foreign dependency of countries. To reduce these negativities, there has been a tendency towards alternative energy sources. The cleanest and most environmentally friendly sources that can be used as alternative energy sources stand out as renewable energy sources. Renewable energy sources, especially solar energy is spreading rapidly with features such as high potential, reliability, ease of use and environmental friendliness. The number of photovoltaic solar power plants is increasing day by day. With the increase in power plants, it has gained importance to examine the parameters affecting the photovoltaic system performance. In this master's thesis study, analysis were made about dusting, shading, and photovoltaic module tilt angle, which affect the system performance, with the measurements to be made in the photovoltaic plant that is currently installed. In the analyzes made, current and voltage measurements were carried out. Output power is calculated for dusty, non-dusted, shaded and shaded panels and the output power differences in different photovoltaic module tilt angles are explained with graphics. These analyzes were analyzed separately for different weather conditions, different shading and different photovoltaic tilt angle situations. In addition, power estimations were made with artificial neural networks in the Matlab program using solar radiation, air temperature, cell temperature, output power, wind speed and direction data from the 3 kW photovoltaic system installed within the scope of the thesis. While making power estimations, different results were tried to be obtained by changing the input data in the artificial neural network. The power estimation differences were investigated using the radiation data from two different locations. The results obtained are explained with graphs for five months and daily according to different weather conditions. Thus, the analysis made about the parameters that contribute to the production of photovoltaic systems and the situations that affect the system performance will guide the new systems to be established. Significant results have been obtained about the production estimation in photovoltaic system designs, the analysis of the variables affecting the estimations and the system structure.