Accuracy Analysis of Power Quality Disturbance Classification Based on Root Sum Square

Title
Accuracy Analysis of Power Quality Disturbance Classification Based on Root Sum Square
Author
Lim, Hong Rok
Alternative Author(s)
임홍록
Advisor(s)
김진오
Issue Date
2019-02
Publisher
한양대학교
Degree
Master
Abstract
The disturbance data of the power quality is essential for protecting the loads connected to the power system. It is important to detect and classify the power quality disturbance accurately in order to solve the fundamental problem of disturbance occurrence by analyzing failure patterns based on the disturbance data of these power quality. In this simulation, there is a process from occurrence of power quality disturbance to data classification. The generated waveform used in the simulation consists of 8 types and they are mathematically modeled. To accurately classify disturbance waveform, 4 types preprocessing methods are used including proposed method. In order to increase the efficiency of these preprocessing, Root Sum Square (RSS) technology is proposed. RSS can be used in conjunction with other preprocessing techniques as well as highlighting the disturbance. Also, RSS has the advantage of being able to be used alone. As a feature extraction, two items are applied, taking into the computational time and storage capacity of data. The extracted feature data can be not only represented through the coordinate plane which can check its distribution of dispersion and cohesion but also the application of RSS can be represented. Therefore, in this paper, the preprocessing and machine learning technique that are suitable for the detection and classification of power quality disturbance based on these numerical results are proposed.
URI
http://dcollection.hanyang.ac.kr/common/orgView/000000108587http://repository.hanyang.ac.kr/handle/20.500.11754/99642
Appears in Collections:
GRADUATE SCHOOL[S](대학원) > ELECTRICAL ENGINEERING(전기공학과) > Theses (Master)
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE