Karaca, Eda; Nuroğlu, Fatih Mehmet

The need to transmit energy in the most effective way arises in parallel with the demand for energy in developing countries. As the amount of transmitted power increases, transmission lines can become overloaded, causing a blackout. This study aims to increase the transmission capacity without experiencing right-of-way problems by converting Turkey's existing high voltage AC systems to hybrid AC/DC systems. With such a design, it is aimed at providing more efficient and continuous transmission by using the high-voltage DC system's low-loss feature at long distances. Using the 400 kV double circuit line structure, the existing electricity transmission system is modelled using real data in the DigSILENT PowerFactory. The electromagnetic fields that will occur due to the use of AC and DC circuits on the same tower in such a transformation situation are analysed with FEMM 4.2, based on the finite element method. The compliance of the electromagnetic fields that will occur with the use of hybrid 400 kV AC and 500 kV DC was determined in the study. In addition, according to the simulation results, it is predicted that in the case of hybrid AC/DC transmission, the current transmission capacity would increase and the losses would decrease.

Altintasi, Cagri; Orguner, Umut; Salor, Ozgul

In this study, the problem of amplitude estimation of (inter)harmonics is investigated. Specifically performance limits are provided for the accuracy of unbiased amplitude estimation algorithms and for the analysis window size required by an algorithm for achieving and maintaining a desired accuracy. Cramer-Rao lower bound (CRLB) is used for the bounds on the accuracy. Unlike the existing literature concentrating on the asymptotic behaviour of the bounds, the behaviour of CRLBs for small window sizes and close frequency components is investigated. In particular, the (worst) amplitude CRLBs are shown to be inversely proportional to the squared frequency differences for small analysis window sizes and close frequency components. For the bounds on analysis window size, the concept of convergence-Time is defined and illustrated. The convergence-Times of unbiased amplitude estimation algorithms are then lower bounded by establishing its relationship with CRLBs. The results and ideas are illustrated using both synthetic signals and field data collected from a power grid. Numerical results show that the proposed performance limits are tight and have strong prediction capabilities for amplitude estimation algorithms having prior knowledge about the existing frequencies in the signal.

Ekici, Sami; Unal, Fatih; Ozleyen, Umit

Various pattern recognition methods have been suggested for estimating high-voltage alternating current transmission line fault location. However, insufficient studies have been conducted on the transmission lines connected to hybrid power generation systems such as wind and solar plants. In this study, the performance of different regression methods was investigated on a hybrid power system. Different faults with random distances on the transmission line were simulated and a fault database created by recording the current and voltage signals of these faults. After normalising this data in the preprocessing phase, it was passed to the digital signal processing stage. By repeating the experiments, 497 different faults were created. Fault types, fault resistances, and fault inception angles were changed randomly in order to obtain similar fault occurrence conditions as in real life by writing a Matlab code. In order to obtain distinctive features, the discrete wavelet transform was used. For training and validation of the dataset, Matlab Regression Learner App (RLA) was employed and the obtained results compared to select the best model. After significant fault simulation, Matern 5/2, a type of Gaussian progress regression model, showed more promising results compared to other RLA models.