Applying Time-Frequency Analysis and Neural Network for U-V Transformer Fault Diagnostic
| Autor: | Wen-Jui Chang, 張文睿 |
|---|---|
| Rok vydání: | 2018 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 107 U-V transformers are used in Taiwan’s power system widely. This issue may cause three phase unbalance and inefficiency in the distribution system. This thesis presents the application of Wavelet Analysis theory and Short Time Discrete Fourier Transform in the power quality analysis of micro-grid and helps grid manager manage the power quality with the tool of 3D graphics and time-frequency data. The technology allows the users to interpret the power quality information with frequency analysis efficiently. This thesis also uses Simulink in Matlab as a tool to construct a small local micro grid, and simulate three phase faults for various kinds of three phase loads. If a fault is detected, the bus voltage and current information will be captured and sent to the diagnostic system for analysis, then the fault will be detected. The mathematical models are based on neural networks and this thesis proposes the use of neural network to classify the fault information after time frequency analysis, and check the accuracy of the system for various cases. The electrical engineers can deal with the unbalanced three phase faults immediately. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|