الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 180 |  |  |
| | | from | 180 | | |
Abstract
Transformers are one of the most important elements in power system. Most of transformers are equipped with protection systems to avoid their damage. As the transformer outage has severed technical and economical consequences for the network, so implementing fast relaying algorithms is a challenge. This thesis introduces a proposed Transformer Internal Fault Model (TIFM) to model the internal winding faults in transformers using ATP in a simple and direct way. The internal faults in all windings are simulated and tested. The results show that the proposed approach is able to represent the internal faults in power transformers accurately.
To avoid the malfunction of the differential relay, alternate improved protection techniques are to be formulated with improved accuracy and high operating speed. This thesis presents an alternative approach using Artificial Neural Networks (ANN) in order to distinguish between inrush currents and internal fault currents as well as estimate the faulted side, classify the fault type and identify the faulted phase. The proposed approach is implemented in MATALB/Script m-files using obtained current waveforms generated by ATP. The simulation results show that the proposed approach for fault diagnosis gives an excellent accuracy, and it is able to detect and classify internal faults of power transformer rapidly, and correctly.