الفهرس 
الغلافOnly 14 pages are availabe for public view
 |  | from | 146 |  |  |
| | | from | 146 | | |
Abstract
Power transformer is an important link in a power system. Internal faults of the transformer may cause a severe damage of transformer. This damage takes a considerable time for repairing. For this reason, faster and accurate protection techniques are required.
The thesis introduces a novel technique based on Discrete Wavelet Transform (DWT) and Adaptive Neuro Fuzzy Inference System (ANFIS) to detect internal faults in power transformers in an early stage, before being developed into more severe faults. The technique is also able to differentiate between internal faults, external faults and inrush condition.
Power transformer simulation was carried out using Alternative Transient Program (ATP) software. The power transformer was modeled accurately as mutually coupled R-L circuits, and its performance during different operating conditions was presented including; normal, internal faults, external faults, and transformer magnetizing inrush current.
The proposed protection technique is based on the combination between Discrete Wavelet Transform (DWT) and Adaptive Neuro Fuzzy Inference System (ANFIS). The DWT is used for extracting the transient features from the differential current. The ANFIS is used as a classifier to discriminate between internal faults and other condition in power transformer. The combination of DWT and ANFIS are set using MATLAB SIMULINK.
The hardware implementation is set using (5 kVA, 220/110 V) laboratory transformer and data acquisition card and Lab- View package.
The obtained results showed that the proposed algorithm is efficient to detect such internal faults and discriminate between it and other disturbances in power transformer, and it provides more secured and dependable results.