Improvement Of Dissolved Gas Analysis Using Artificial Intelligence Approach

Amran, Ahmad Haziq (2018) Improvement Of Dissolved Gas Analysis Using Artificial Intelligence Approach. Project Report. UTeM, Melaka, Malaysia. (Submitted)

Text (24 Pages) Improvement Of Dissolved Gas Analysis Using Artificial Intelligence Approach.pdf Download (417kB)

Abstract

This project is to design a robust and reliable intelligent diagnostic method to detect and predict incipient faults in transformer. Transformer is one of the most important components in the power system network. Major fault in these transformers can cause extensive damage which is not only disturbing other features electricity supply, instead causing huge losses. In the transformer, insulation material and faulty equipment will result in the release of gas, hence can be attributed to some kind of electrical fault such as corona, pyrolysis and arcing. The resulting gas generation rate can indicate the severity of the offence and the information obtained can be very beneficial in any preventive maintenance program. By using any of the preventive maintenance programs, the identity of gas is very useful to determine that faults. The key gas considerations for evaluation are hydrogen (H2), methane (CH4), ethane (C2H6), ethylene (C2H4), acetylene (C2H2), carbon monoxide (CO) and carbon dioxide (CO2). Thus, interpretation of dissolved gas analysis (DGA) is used as the preventive maintenace program to detect the incipient faults. To study in DGA related to incipient fault inside power transformer, several interpretation methods for DGA will be discussed. The interpretation methods are Key Gas Method (KGM), Doernenburg Ratio Method (DRM), Rogers Ratio Method (RRM), IEC Ratio Method (IRM) and Duval Triangle Method (DTM). In order to automate this program, the technique of artificial inteligence by using MATLAB software is developed in this study. Artificial intelligence method is selected because of its ability in storing knowledge and their functions to make decision.

Actions (login required)

View Item