Bir bölgenin tek bant GSM RF enerji hasatlama potansiyelinin modellenmesi

In this thesis, it is aimed to model a region with machine learning to reveal the single band GSM RF energy harvesting potential and capacity. Since it is possible to harvest energy from many different ambient energies, similar studies have been carried out for many other sources such as sun, heat, light, piezoelectricity. In the thesis study, a sufficient number of basic level RF Energy Harvesting circuits were designed in the GSM 900MHz single band frequency band and a data set was created by measuring the amount of energy stored on the capacitor of this circuit on a daily basis. After observing the energy levels stored in the RF energy harvesting circuit for different time periods, the energy potential of the region was modeled with the help of linear regression, xgboost, arima and fb prophet algorithms, which are frequently used in machine learning algorithms, and a conclusion was reached about the potential of the region. Comparison of the performances of estimation algorithms was evaluated based on the MSE (Mean Square Error), MAE (Mean Absolute Error), RMSE (Root Mean Square Error), MAPE (Mean Absolute Percent Error) statistics, which are frequently used in machine learning.