Güneş fotovoltaik sistemlerinin performansı için güvenilirlik modellemesi

Solar energy is one of the most widely used renewable energy sources. Photovoltaic systems directly convert solar energy into electricity with no carbon dioxide emission or any other air pollutants. The power generated by a photovoltaic system depends on the characteristics of solar irradiation and weather conditions. This stochastic nature of power systems prompts the researchers to use probabilistic and statistical techniques. In this thesis, we consider the distribution function of the power generated by a photovoltaic system to make predictions about its characteristics. We model the mean power generated by a photovoltaic system. We also consider that photovoltaic modules may have multistate working conditions and different performance levels depending on solar radiation. In this concept, we present a model for solar power systems with PV modules having various levels of operational performance and we develop a reliability model for the system's power regarding the threshold value that is the minimum required total performance level for the system. This model reflects the performance levels and the working probabilities of PV modules. The problem is evaluated under different conditions regarding the dependency of multistate PV modules. In addition, we provide the optimum number of photovoltaic modules that minimizes the total cost based on the level of required total power production. For further analyses, we give real data applications to estimate the characteristics of the power produced by the solar plant for a specific location in Izmir, Turkey. As the software programming tools, R v.1.2.5033 and Mathematica v.11.3 are used for the computations.