Analysis of power quality distortions in electrical energy systems with intelligent methods

The development and differentiation of production technology has led to diversity in loads in transmission/distribution systems. One of the biggest effects of this diversity is the increase in non-linear loads in the system with the increase in semiconductor technology. Nonlinear loads are defined as loads whose current and voltage characteristics are not pure sinusoidal. The main non-linear loads are power electronic components, transformers and high-power induction motors. Harmonics cause additional energy losses, heating and damage to the insulation in the system. Therefore, it is important to eliminate harmonics before they occur. In this thesis, it is aimed to prevent the damages that may occur in the distribution system by forecasting the harmonics with intelligent methods. In this context, an LSTM based estimation algorithm has been developed by using real system power quality data from Bandırma Organized Industrial Zone. By using real data from a physical system in the thesis, parameters that are likely to be neglected in simulation studies are also included in the calculation. The results obtained were discussed over graphics, and it was observed that the forecasting performance was quite high.