Estimation of the effects of trash rack on efficiency in riverside type power power plants using machine learning

The need for electricity and therefore electricity consumption in the world and in our country is increasing every year. Fossil energy sources used for energy needs from past to present cause increased environmental pollution and global warming. Increasing global warming negatively affects the basic living resources required for living things to survive. In order for the world to be a livable place in the future, energy must be produced in the most harmless way. In order to prevent this situation, it is necessary to reduce the use of fossil energy resources used for electricity production and cause global warming and increase the use of renewable energy resources instead. Hydraulic energy resources, which are among the renewable energy resources, are of great importance among energy resources as their use becomes widespread in the world. For this reason, it is also important to produce energy with high efficiency in Hydraulic Power Plants. In this context, in this thesis, the efficiencies of Hydraulic Power Plants are analyzed in the Karkamış HEPP sample, which is a river type power plant. The analysis was made separately for each unit and the effect of trash rack pollution above head flow rate and production values was investigated. As a result of the analysis carried out in this context, the efficiency-flowrate relationship of trash rack pollution is revealed. Head losses due to trash rack pollution are examined. Thus, the scope and definition of the problem was made. Data was received regularly from the Karkamış HEPP SCADA system for two years. These data amounts have resulted in matrices exceeding millions of rows. The data in question was arranged according to the same time period and turned into a separate data set for each of the five units. After filtering the data sets, machine learning and trash rack pollution prediction studies are carried out by using the study data in the MATLAB program. After filtering the data sets, machine learning and trash rack pollution guessworks are carried out by using the study data in the MATLAB program. Machine learning was carried out with regression analysis in the MATLAB program. Tree models, support vector machine models, ensemble learning models and neural network models in the linear regression analysis of the MATLAB program were used in machine learning and guessworks were carried out. The results of the evaluation of the study are analyzed graphically and through R, RMSE and MAE values. The work done for each unit is interpreted. Suggestions have been made to prevent trash rack pollution and to increase efficiency by operating the units in the power plant according to the trash rack pollution forecast program. Suggestions have been made to prevent trash rack pollution and to increase efficiency by operating the units in the power plant according to the trash rack pollution forecast program.