Load Classification Using Self-Organizing Map (SOM) And Correlation Analysis For Optimization

Azman, Fatin Nabila (2018) Load Classification Using Self-Organizing Map (SOM) And Correlation Analysis For Optimization. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (24 Pages) Load Classification Using Self-Organizing Map (SOM) And Correlation Analysis For Optimization.pdf - Submitted Version Download (533kB)

Abstract

The Malaysian electricity sector was regulated to power supply and has seen as the independent power producers (IPPs).The natural vulnerabilities related with load classification end up more intense when it actually required providing a period characterized by quick and energetic changes for priceless measurement to the decision-making process.In this research,a neural network approach is proposed for load classification in Malaysia. This research presents the clustering load using an approach based on the combination of Self-Organizing Maps (SOM) and Correlation Analysis for optimization.The main purpose in this project is to understand the ability of correlation analysis and SOM in load classification,and to relate and train the load data via correlation analysis and SOM method using selected features.This project focused on clustering data in January 2016 until early February 2016.The data from previous year was trained to cluster the data for the future planning.The relation of load data was correlated and cluster using calculated correlation analysis and SOM Toolbox in MATLAB software.After that,the map size,training time, quantization error and topographic error is compute to determine the performance of the result.

Actions (login required)

View Item