Self Organizing Map Technique For Transformers Condition Monitoring From SFRA Result

Ahmad Nor Syahid, Hassan (2013) Self Organizing Map Technique For Transformers Condition Monitoring From SFRA Result. Project Report. UTeM, Melaka, Malaysia. (Submitted)

	PDF (24 Pages) Self_Organizing_Map_Technique_For_Transformers__Condition_Monitoring_From_SFRA_Result_-_24_Pages.pdf - Submitted Version Download (530kB)
	PDF (Full Text) Self_Organizing_Map_Technique_For_Transformers__Condition_Monitoring_From_SFRA_Result_-_Full_Text.pdf - Submitted Version Restricted to Registered users only Download (2MB)

Abstract

Self organizing map (SOM) is the main method proposed for this project where the purpose was using it to determine and also diagnosing the data taken from various condition monitoring tests in the real distribution transformers application. SOM is the tool that uses to decide and collect data by give a train, where the sum of data will give as a reference to the SOM and then it will explore the data and also give a conclusion result about the data. This tool is an unsupervised neural network where it has their own algorithm which is consist vector quantization and vector projection. When have a lot of data using SOM it will make it their decision using a reference before where it supervise the new data and conclude it in what the value as reference decided before. In this project, by applying a data from Frequency Response Analysis (FRA) apply it to SOM tools to obtain the result in topographic map. This project had focus on using toolbox in Matlab with perform with their algorithm and create the data set to put as input using m file. This research work had achieved their objective and successfully gives the good result according the Self Organizing Map technique.

Actions (login required)

View Item