Reconfiguration method and optimization based analysis using image processing

Energy requirement for human being has been increasing with the development of technology. A variety of energy sources has been started to investigate in order to supply this increasing demand. The energy sources have been divided two types of energy sources as renewable and non-renewable. Solar energy from renewable energy sources has become very popular lately due to the reasons such as being easy to get and low cost. In order to utilize of this energy, photovoltaic panels (PV), which are constructions that convert solar energy into electricity, have been produced. Thus, great amount of energy has been obtained. However, various problems are encountered when energy is generated from PV arrays. Full or partial shadows cause the maximum power tracking algorithm used in PV array to produce incorrect values. Thus, the system is worked inefficiently and prevents obtaining maximum energy from the system. A lot of studies have been done in order to solve the problem. One of the studies for increasing the energy obtained from PV arrays is the reconfiguration algorithms. With these algorithms, real-time processing is performed to obtain the most suitable panel connections and the optimum panel layout is created. In this manner, it is intended to minimize the effects of negative states that occurred. In this thesis study, it is aimed to minimize the damage of this problem by seeking a solution to the problems of shadowing. For this purpose, using image processing a genetic algorithm based reconfiguration approach has been developed. The study consists of three steps; image processing, genetic algorithm and control and decision module. The real time images obtained from the panels are taken and sent to the image processing module. The obtained images are processed into the genetic algorithm module to create optimum panel layouts. The final stage is sent to the control and decision module. Optimum panel layout is obtained by selecting one of the most suitable panel layout created in this module. The studies developed within the scope of this thesis were supported by the research project of TÜBİTAK 1001; the project number 112E214.