Pyeong-Shik Ji

Diagnosis of Power Transformer using Fuzzy Clustering and Radial Basis Function Neural Network

Diagnosis techniques based on the dissolved gas analysis (DGA) have been developed to detect incipient faults in power transformer. There are various methods based on DGA such as IEC, Roger, Dornenburg, and etc. However, these methods have been applied on the different problems with different standards. Also, it is difficult to achieve the diagnosis with accuracy by DGA without experienced experts. In order to resolve theses drawbacks, this paper proposes a novel diagnosis method using fuzzy clustering and a radial basis neural network (RBFNN). In neural network, fuzzy clustering is effective for selecting the efficient training data and reducing learning process time. After fuzzy clustering, the RBF neural network is developed to analysis and diagnosis the state of transformer. The proposed method measures the possibility and degree of aging as well as the faults occurred in transformer. To demonstrate the validity of proposed method, various experiments are performed and their result is presented.

A load modeling using ANN for power system analysis

A load model is very important to improve the accuracy of stability analysis and load flow study in power systems. A power system bus is composed of various loads, and these loads have different power consumptions to voltage/frequency changes. Thus, the effect of voltage/frequency changes must be considered in load modeling. In this research, an ANN was used to construct a component load model for more accurate load modeling. A typical residential load was selected, and experimented on voltage/frequency changes. Acquired data used to construct component ANN models, and an aggregation method of the component load model are presented based on component load model and composition rate. Furthermore, the transformation method used in traditional power system analysis software is also presented.

Fault Diagnosis of Power Transformer Using SVM and FCM

In this study, we are concerned with fault diagnosis of power transformer. The objective is to explore the use of some advanced techniques such as SVM and FCM and quantify their effectiveness when dealing with dissolved gases extracted from power transformers. The proposed fault diagnosis system consists of data acquisition, fault/normal diagnosis, identification of fault and analysis of aging degree parts. In data acquisition part, concentrated gases are extracted from transformer for data gas analysis. In fault/normal diagnosis part, SVM is performed to separate normal state from fault types. The determination of fault type is executed by multi-class SVM in identification part. Although the inputted data is normal state, the analysis of aging degree is performed by considering the distance measure calculated by comparing with reference model constructed by FCM and input data. Our approach makes it possible to measure the possibility and degree of aging in normal transformer as well as the identification of faults in abnormal transformer. As the simulation results to verify the effectiveness, the proposed method showed more improved classification results than conventional methods.

Load characteristics identification using artificial neural network and transient stability analysis

The modeling of load characteristics is a difficult problem because of uncertainty of the load. This research uses artificial neural networks which can approximate the nonlinear problem to represent load characteristics. After the selection of a typical load, active and reactive power for the variation of voltage and frequency is obtained from experiments. On the basis of obtained data, the load model represented by a neural network is acquired Then the propriety is submitted by case studies.

Development of Fault Diagnosis Algorithm using Correlation Analysis and ELM

It is difficult to establish accurate modeling of PV power system because of various uncertainty. However, it is important work to modeling of PV for fault diagnosis. This paper proposes modeling and fault diagnosis method using correlation analysis and ELM(Extreme Learning Machine). Rather than using total data, we select optimal time interval with higher corelation between PV power and solar irradiation. Also, we use average value during 60 minute to avoid rapid variation of PV power. To show the effectiveness of the proposed method, we performed various experiments by dataset.

Development of Peak Power Demand Forecasting Model for Special-Day using ELM

Abstract - With the improvement of living standards and economic development, electricity consumption continues to grow. The electricity is a special energy which is hard to store, so its supply must be consistent with the demand. The objective of electricity demand forecasting is to make best use of electricity energy and provide balance between supply and demand. Hence, it is very important work to forecast electricity demand with higher precision. So, various forecasting methods have been developed. They can be divided into five broad categories such as time series models, regression based model, artificial intelligence techniques and fuzzy logic method without considering special-day effects. Electricity demand patterns on holidays can be often idiosyncratic and cause significant forecasting errors. Such effects are known as special-day effects and are recognized as an important issue in determining electricity demand data. In this research, we developed the power demand forecasting method using ELM(Extreme Learning Machine) for special day, particularly, lunar new year and Chuseok holiday.Key Words : ELM, Forecasting model, Power demand, Special-day *Dept. of Electrical Engineering, Korea National University of Transportation, Korea†Corresponding Author : Dept. of Electrical Engineering Daeduk College, Korea E-mail : jylim@ddc.ac.kr접수일자 : 2015년 5월 3일최종완료 : 2015년 5월 20일

Development of Daily Peak Power Demand Forecasting Algorithm using ELM

Due to the increase of power consumption, it is difficult to construct an accurate prediction model for daily peak power demand. It is very important work to know power demand in next day to manage and control power system. In this research, we develop a daily peak power demand prediction method based on Extreme Learning Machine(ELM) with fast learning procedure. Using data sets between 2006 and 2010 in Korea, the proposed method has been intensively tested. As the prediction results, we confirm that the proposed method makes it possible to effective estimate daily peak power demand than conventional methods.

Fault Diagnosis of Power Transformer Using Support Vector Machine

For the fault diagnosis of power transformer, we develop a diagnosis algorithm based on support vector machine. The proposed fault diagnosis system consists of data acquisition, fault/normal diagnosis, and identification of fault. In data acquisition part, concentrated gases are extracted from transformer for data gas analysis. In fault/normal diagnosis part, KEPCO based decision rule is performed to separate normal state from fault types. The determination of fault type is executed by multi-class SVM in identification part. As the simulation results to verify the effectiveness, the proposed method showed more improved classification results than conventional methods.

Modeling of Photovoltaic Power Systems using Clustering Algorithm and Modular Networks

The real-world problems usually show nonlinear and multi-variate characteristics, so it is difficult to establish concrete mathematical models for them. Thus, it is common to practice data-driven modeling techniques in these cases. Among them, most widely adopted techniques are regression model and intelligent model such as neural networks. Regression model has drawback showing lower performance when much non-linearity exists between input and output data. Intelligent model has been shown its superiority to the linear model due to ability capable of effectively estimate desired output in cases of both linear and nonlinear problem. This paper proposes modeling method of daily photovoltaic power systems using ELM(Extreme Learning Machine) based modular networks. The proposed method uses sub-model by fuzzy clustering rather than using a single model. Each sub-model is implemented by ELM. To show the effectiveness of the proposed method, we performed various experiments by dataset acquired during 2014 in real-plant.

Design of Long Span Overhead Transmission Line using Special High-tension Wire

Recently, power demand has been increasing every year according to variation of electrical equipments and temperature rise in summer season. So, much more overhead line is being demanded to copy with increasing power demand and operate reliable power system. This paper analysis the characteristics of long span overhead transmission line using special high-tension wire in such as a safety factor, coefficient of elasticity, and the coefficient of linear expansion. Based on the analysis, we proposed the effectiveness of special high-tension wire having much more advantages with respect to height of steel tower and dip compared with conventional ACSR in long span overhead transmission line.