TY  - THES
T1  - A dissolved gas analysis-based failure prediction model for power transformers using artificial neural networks
A1  - Pati, Eduardo M.
LA  - English
YR  - 2004
UL  - https://ds.mainlib.upd.edu.ph/Record/UP-99796217607612500
AB  - This paper presents a three-step artificial neural network (ANN) approach to predict power transformer failures based on dissolved gas-in-oil analysis (DGA). The purpose is to use ANN-based detectors to initially classifiy the condition of a power transformer. Once the abnormal condition is established, another sets of ANN detectors are then employed to determine the type of fault and possible involvement of cellulose. All of these can be accomplished accurately in a short period of time and solve the problem on over reliance of human experts to evaluate the condition of a power transformer using conventional DGA methods. The developed ANN model was tested on various cases of power transformers. The results indicate that compared to conventional DGA methods, the developed ANN model has been able to accurately identify the true condition of the power transformers in a short period of time.
CN  - LG 995 2004 E64 P38
KW  - Neural networks (Computer science).
KW  - Gases : Analysis.
KW  - Electric discharges : Detection : Computer simulation.
KW  - Electric power failures : Computer simulation.
KW  - Electric transformers.
ER  -