Investigation of distance protection relay operation in mixed transmission lines and improving operation performans

In this thesis, high voltage 154 kV transmission high impedance distance protection relays used in the protection of the lines short circuit, power transformer grounding system of short-circuit failure of the differences and mixed transmission lines in overhead lines or underground cable line in the resulting short circuit fault and work on identifying the fault location is presented. Intelligent system design is recommended using image processing technique and artificial neural networks (ANN) to accurately locate short circuit faults that may occur in transmission lines. Energy transmission line model PSCAD ™ / EMTDC ™ simulation program created in a short circuit fault distance protection relays in R-X on the impedance change in the impedance diagram recorded image data sets have been created. ANN model was selected images feature extraction and image processing techniques have been focal points as an introduction to using different neural network model and the best estimate that fault location. With the prepared data set, various network types, learning functions, training functions and performance functions were tested and the most suitable network structure was determined. Single phase to earth to determine fault location for failure problem feedforward backpropagation, cascade feedforward backpropagation and Elman backpropagation three different types of network with Levenberg-Marquardt (LM), Conjugate Gradient with Powell / Beale Restarts (CGB) One Step Secant (OSS), Variable rate Backpropagation Learning (GDX) and Scaled Conjugate Gradient (SCG) five different educational training error rate function of the results obtained using the comparison evaluation was performed. In the thesis, the images we obtained in the distance transmission relay in the long transmission lines, in the case of high impedance short circuit faults, if the grounding system is different and in the mixed transmission lines, the problem of determining the fault location is made by using the image processing technique and the realistic and high estimation values are shown with charts.