Development of new methods for detecting disturbance in power systems based on amplitude and frequency evaluation

Increasing competition in the energy sector and users' demand for better quality energy have made power quality a priority issue in electrical networks. In order for the energy used in electrical networks to be of high quality, it must be continuous, at a certain voltage level, with a fixed frequency and amplitude. However, these are not normally provided in full. In electrical power systems, voltage and current waveforms deviate from linearity due to non-linear elements in the voltage-current characteristic. Power quality problems have also increased with the increase in nonlinear loads, power electronic circuit elements and circuits in today's modern power systems. Harmonics are one of the main factors affecting power quality. Elements such as nonlinear loads, rectifiers, arc furnaces and inverters in power transmission lines produce harmonics. The effects of harmonics on electrical power systems are; It creates many undesirable situations such as overheating, voltage drops, additional losses, resonance events, dielectric stress, malfunction of protection and control measurement systems. In electrical systems, due to the operation of non-linear loads, especially Electric Arc Furnaces, it is of great importance to detect the current and voltage source at common connection points and to determine how many components are contributed to the electrical lines. Therefore, detecting, measuring, estimating and suppressing harmonics is extremely important. Accurate prediction and prevention of harmonics in terms of amplitude, frequency and phase improves power quality and efficiency. With the development of artificial intelligence and smart systems, smart methods as well as Fourier transform-based algorithms are used for the estimation of harmonics. In this thesis study, hybrid methods consisting of the least squares method and many meta-heuristic algorithms have been tried and developed for the amplitude and phase estimation of harmonics. While the amplitude of harmonics was calculated by the least squares method, phase angles were estimated by many optimization algorithms. The success of the proposed algorithms has been tested using the test signals suggested in the literature and the real data set. Harmonics in powers have been a frequently studied research area in recent years. This thesis study started with the first screening using PSO (Particle Swarm Optimization) and Genetic Algorithm on harmonic signals in the literature, and then many versatile methods were tried. Methods have been tried at different frequencies using EOA (Election Based Optimization), GWO (Grey Wolf Optimizer) and GOA (Grasshopper Optimization). Afterwards, harmonic estimations were made with AVOA (African Vulture Optimization Algorithm), ARO (Artificial Rabbit Optimization), SWO (Spider Wasp Optimization), MGO (Mountain Gazelle Optimization) and AO (Aquila Optimization) algorithms developed in recent years. In the analyses, several synthetic data samples frequently used in the literature were studied. The harmonic amplitudes of this signal were determined by the Least Squares (LS) method, and the phase angles were estimated using the relevant metaheuristic algorithms. Harmonic estimation has been successfully achieved with the AO algorithm and other metaheuristic algorithms, which give the best results in the synthetic data sample on the real data set File wave14a.xls in the IEEE 1159.2 Working Group.The developed algorithms were tested in the MATLAB software environment and the results were discussed by comparing them in many error metrics with similar studies and all the algorithms that we made validity analysis for. The results obtained showed that all method estimates examined provided accurate and reliable harmonic detection even in noiseless and noisy conditions. In the detection of harmonics, the hybrid use of the AO algorithm and the LS method provided the most accurate and consistent results. The proposed AO-LS method has been validated by applying it on different test data sets. In addition, it has been observed that the harmonic detection performance is further improved with the SALSAO (SA-LS-AO) method, which is obtained by combining the AO-LS method with the Simulated Annealing (SA) algorithm in a hybrid way. The hybrid method SALSAO proposed in this thesis was first tested in the detection of harmonic distortions, and after obtaining good results, it was also tested in the analysis of other types of distortions in power systems and successful results were obtained. The performance of the method was evaluated on both synthetic test data and real data and it was found that it provides high accuracy and reliability even in noisy and noiseless conditions. The findings revealed that the SALSAO algorithm is an effective approach in solving power quality problems.