• KIPS Transactions on Computer and Communication Systems
• /
• 제5권10호
• /
• pp.293-302
• /
• 2016

Since the electricity is produced and consumed simultaneously, predicting the electric load and securing affordable electric power are necessary for reliable electric power supply. In particular, a university campus is one of the highest power consuming institutions and tends to have a wide variation of electric load depending on time and environment. For these reasons, an accurate electric load forecasting method that can predict power consumption in real-time is required for efficient power supply and management. Even though various influencing factors of power consumption have been discovered for the educational institutions by analyzing power consumption patterns and usage cases, further studies are required for the quantitative prediction of electric load. In this paper, we build an electric load forecasting model by implementing and evaluating various machine learning algorithms. To do that, we consider three building clusters in a campus and collect their power consumption every 15 minutes for more than one year. In the preprocessing, features are represented by considering periodic characteristic of the data and principal component analysis is performed for the features. In order to train the electric load forecasting model, we employ both artificial neural network and support vector machine. We evaluate the prediction performance of each forecasting model by 5-fold cross-validation and compare the prediction result to real electric load.

• Journal of the Korean Society for Advanced Composite Structures
• /
• 제3권2호
• /
• pp.19-27
• /
• 2012

Structural insulated panels, which are structurally performed panels consisting of a plastic insulation bonded between two structural panel facings are one of emerging products with a viewpoint of its energy and construction efficiencies. These components are applicable to fabricated wood structures. By now, there are few technical documents regulated structural performance and engineering criteria in domestic market. This study was conducted to suggest fundamental reports such as racking resistance, axial capacity, transverse load capacity, and lintel load capacity for SIPs. Test results showed that maximum load was 44.3kN, allowable load was 14.7kN for racking resistance, and that maximum load was 137.6kN, allowable load was 37.4kN/m for axial compression capacity. For transverse load capacity, test results showed $10.3kN/m^2$ of maximum load, $3.4kN/m^2$ of allowable load. For lintel load capacity for SIPs dependent to lengths, allowable loads were 20.4kN for 600mm long lintel, 23.9kN for 1,200mm long lintel, 19.3kN for 1,800mm long lintel, and 2,400mm long lintel had 14.1kN of allowable load. In the near future, when the allowable load for wall application is established, SIPs is considered to substitute the existent post-and-lintel construction to bearing wall structure.

• Korean Journal of Remote Sensing
• /
• 제23권5호
• /
• pp.393-400
• /
• 2007

We present in this paper a novel power load prediction method using temporal pattern mining from AMR(Automatic Meter Reading) data. Since the power load patterns have time-varying characteristic and very different patterns according to the hour, time, day and week and so on, it gives rise to the uninformative results if only traditional data mining is used. Also, research on data mining for analyzing electric load patterns focused on cluster analysis and classification methods. However despite the usefulness of rules that include temporal dimension and the fact that the AMR data has temporal attribute, the above methods were limited in static pattern extraction and did not consider temporal attributes. Therefore, we propose a new classification method for predicting power load patterns. The main tasks include clustering method and temporal classification method. Cluster analysis is used to create load pattern classes and the representative load profiles for each class. Next, the classification method uses representative load profiles to build a classifier able to assign different load patterns to the existing classes. The proposed classification method is the Calendar-based temporal mining and it discovers electric load patterns in multiple time granularities. Lastly, we show that the proposed method used AMR data and discovered more interest patterns.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• 제50권1호
• /
• pp.20-27
• /
• 2001

In this paper, we propose a method for estimating the peak load of pole-transformer in summer season considering the degree of cooling load possession in customer. The cooling load of customer is selected as the most reliable parameter of peak load in summer season. The proposed estimation method is restricted to the aspect of load management for pole-transformer. The main concept of proposed method is that the error of peak load estimation using load regression equation reduces with considering the degree of cooling load possession in customer. We propose an index for estimation of cooling load possession in each customer. The proposed index is defined as cooling load possession in customer (CLPC) and obtained from the increment of monthly electric energy. The membership function for deciding the uncertainty of cooling load possession in customer is used. The database of pole-transformer in Korea Electric Power Corporation (KEPCO) is used for case studies. Through the case studies, we verify that the proposed method reduces the error of peak load estimation than the conventional method in domestic.

• Journal of Ocean Engineering and Technology
• /
• 제33권5호
• /
• pp.447-453
• /
• 2019

Recently, environmental regulations have been strengthened for SOX, NOX, and CO2, which are ship exhaust gases. In addition, according to the 4th Industrial Revolution, research on autonomous ship technology has become active and interest in electric propulsion systems is increasing. This paper analyzes the power load characteristics of an electric propulsion ship, which is the basic technology for an autonomous ship, in terms of energy management. For the load analysis, data were collected for a 6,800 TEU container ship with a mechanical propulsion system, and the propulsion load was converted to an electric power load and clustered according to the characteristics using a SOM (Self-Organizing Map). As a result of the load analysis, it was confirmed that the load characteristics of the ship could be explained by the operation mode of the ship.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• 제64권7호
• /
• pp.971-977
• /
• 2015

The electric propulsion system requires more reliability and safety than the conventional propulsion system because any sudden changes of electric system would bring tremendous effects on the ship's safety and propulsion. So it is very important to consider the potential transient effects. This paper discusses one of the worst electric accident. That is, one or two of generators are out of service in normal seagoing condition. And the appropriate measures are simulated in order to prevent the frequency decline that can bring the other generator's tripping. In addition, the relation between the transient effects and the major factors(inertia of generator/motors, governor's drooping characteristic and response speed) are also identified using the ETAP software.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• 제56권1호
• /
• pp.6-12
• /
• 2007

In this paper, an robust adaptive backstepping controller is proposed for the speed control of separately excited DC motor in pure electric vehicles. A general electric drive train of PEV is conceptually rearrange to major subsystems as electric propulsion, energy source, and auxiliary subsystem and the load torque is modeled by considering the aerodynamic, rolling resistance and grading resistance. Armature and field resistance, damping coefficient and load torque are considered as uncertainties and noise generated at applying load torque to motor is also considered. It shows that the backstepping algorithm can be used to solve the problems of nonlinear system very well and robust controller can be designed without the variation of adaptive law. Simulation results are provided to demonstrate the effectiveness of the proposed controller.

• Journal of international Conference on Electrical Machines and Systems
• /
• 제2권1호
• /
• pp.104-110
• /
• 2013

In this paper an application of centralized pitch angle controller for fixed speed wind turbines based wind farm to mitigate load frequency fluctuation is presented. Reference signal for the pitch angle of each wind turbine is calculated by using proposed centralized control system based on wind speed information. The wind farm in the model system is connected to a multi machine power system which is composed of 4 synchronous generators and a load. Simulation analyses have been carried out to investigate the performance of the controller using real wind speed data. It is concluded that the load frequency of the system can be controlled smoothly.

• Proceedings of the KIEE Conference
• /
• 대한전기학회 1996년도 추계학술대회 논문집 학회본부
• /
• pp.163-165
• /
• 1996

In this paper, we are proposed development of electric energy method and load curve. A daily electric energy is forecasted using artificial neural network. The load curve is obtained by combining forecasted electric energy and typical daily load patterns which are classified using KSOM and Fuzzy system. As a result, we know that we could get more accurate results and easier application than the results from based on the hourly historical data.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• 제67권1호
• /
• pp.96-102
• /
• 2018

For efficient operating strategy of electric power system, forecasting of daily peak electric load is an important but difficult problem. Therefore a daily peak electric load forecasting system using a neural network and fuzzy system is presented in this paper. First, original peak load data is interpolated in order to overcome the shortage of data for effective prediction. Next, the prediction of peak load using these interpolated data as input is performed in parallel by a neural network predictor and a fuzzy predictor. The neural network predictor shows better performance at drastic change of peak load, while the fuzzy predictor yields better prediction results in gradual changes. Finally, the superior one of two predictors is selected by the rules based on rough sets at every prediction time. To verify the effectiveness of the proposed method, the computer simulation is performed on peak load data in 2015 provided by KPX.