Power transformer hybrid dissolved gases analysis (DGA) faults interpretation using fuzzy logic
Loading...
Date
2019-06
Authors
Muhammad Syazreen Bin Mohd Daud
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This project is to develop an integrated Dissolved Gases Analysis (DGA)
diagnosis methods and a Combinational (Hybrid) system to detect and predict incipient
faults inside power transformer. Power transformers are one of the most used and
expensive equipment in many substations of electrical system. This fact justifies the
application of predictive maintenance techniques to run diagnosis with the objective to
minimize possible failures and to increase the trustworthiness of the system. In
transformer, degradation of insulation materials and incipient faults in the equipment will
result in the release of hydrocarbon gases. These incipient faults are such as overheating,
electrical discharges activities, pyrolysis, and arching. The resulting gases generation rate
that used to indicate the severity of incipient faults inside transformer. This information
obtained can be very beneficial in any preventive maintenance program. An expert system
was developed in this research to classify and diagnose incipient faults presence inside
transformer using Fuzzy Logic controller system through Simulink block diagram in
MATLAB. The key gases used for the diagnosis are hydrogen (H2), methane (CH4),
ethane (C2H6), ethylene (C2H4), and acetylene (C2H2). The existing interpretation of
Dissolved Gas Analysis (DGA) methods listed in IEC 60599 is used to classify the fault
gases in this project. These methods are Rogers Ratio method, IEC Ratio method, Key
Gas method and Duval Triangle method. In order to have an automatic diagnosis system,
intelligent system for each method was develop using MATLAB software by
implementing Fuzzy Logic. Fuzzy Logic is selected due to its ability in storing knowledge
and of their functions to make decision. The research found by using Fuzzy Logic to
diagnose all four individual DGA interpretation methods is less accuracy due to high
number of non-predictable condition. So, by developing Hybrid system combination of
all four DGA interpretation methods will give higher accuracy to the system for diagnosis
the final incipient faults presence inside power transformer.