Akmaz, DuzgunMamis, Mehmet SalihArkan, MuslumTagluk, Mehmet Emin2024-08-042024-08-0420180378-77961873-2046https://doi.org/10.1016/j.epsr.2017.09.019https://hdl.handle.net/11616/97981In this research, a new approach was proposed for determining the fault location in transmission lines. Traveling wave frequencies and an extreme learning machine (ELM) were used to determine fault location. Transient signals in the time domain were transformed to the frequency domain using the fast Fourier transform (FFT) and the traveling wave frequencies were detected from the transient frequency spectrum. In order to detect the location of fault, traveling wave frequency was used initially to predict the fault location. The prediction of this fault location was tested for many different fault conditions and was found to be adversely affected by only the source inductance value. This is due to the negative effect of source inductance on wave velocity. Regression feature of ELM was used in order to improve the prediction of fault location and to minimize the negative effect of source inductance. For ELM regression training, values of the fault distance estimated from the traveling-wave frequencies and the source inductance values were used as ELM input data, and the actual distance values were used as ELM output data. After ELM regression training, ELM predicted a new fault location using the input data. The Alternative Transients Program (ATP/EMTP) was used to model J. Marti frequency dependent line model, and the MATLAB program was used to perform fault-detection algorithms. Simulation results show that the proposed method is very successful against many variables such as different fault resistances, source inductances, transmission line characteristics, transmission line lengths. (C) 2017 Elsevier B.V. All rights reserved.eninfo:eu-repo/semantics/closedAccessTransmission linesFault-location estimationFast Fourier transformTraveling wave frequenciesExtreme learning machineTransmission line fault location using traveling wave frequencies and extreme learning machineArticle1551710.1016/j.epsr.2017.09.0192-s2.0-85030091064Q1WOS:000419410300001Q2