• 전기의세계
• /
• v.53 no.7
• /
• pp.31-35
• /
• 2004

본 고에서는 현재 개발 및 필드테스트되고 있는 전력품질 보상기기인 STATCOM(Static Compensator), DVR(Dynamic Voltage Restorer), 고속전력절체스위치 SSTS(Solid State Transfer Switch)들이 국내 22.9kV급 배전계통에 도입되어 운용될 경우, 배전계통의 전력품질상의 외란, 즉 voltage sag, voltage swell, interruption등에 대해서도 부하기기가 문제없이 잘 동작할 수 있어야 하는 전력품질향상기기의 성능(전력품질향상기준)에 대한 평가방법 에 대하여 기술하고자 한다. 먼저, 전력품질향상기기인 DVR, D-STATCOM, SSTS들이 만족해야할 전력품질향상기준을 국내 전력 품질 실태를 고려하여 제시하고, 또한, 이들 전력품질향상기기에 대한 국내 배전계통적용에 적합한 성능평가방안을 제안/소개하기로 한다. (중략)

• Proceedings of the KIEE Conference
• /
• 2003.07d
• /
• pp.2657-2659
• /
• 2003

전력선 통신(Power-Line Communication)에 있어서 패널의 임펄스 응답은 송수신 신호의 기본적인 전달 특성이라는 점에서 매우 중요한 채널 특성이며, 채널 구성 요소들의 물리적 구조 및 각종 부하 기기들에 따라 다양하게 변화하기 때문에, 적절한 측정과 분석 및 대응이 요구된다. 이에 본 논문에서는 채널 임펄스 응답 측정에 대한 광대역 측정 방식과 협대역 측정 방식을 검토하고, 각 방식을 테스트 베드 상에서 구현하여 임펄스 응답 측정 결과를 상호 비교하고, 그 신뢰성을 확인하였다. 또, 실제 전력선 채널 환경에서 PN 시퀀스를 이용한 광대역 측정 방식의 장점을 고려하여, 국내 일반적인 저층 아파트를 대상으로 임펄스 응답 측정을 수행하고 결과를 분석하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.10b
• /
• pp.302-307
• /
• 2007

RFID를 기반으로 유비쿼터스 환경의 응용 서비스를 지원하는 미들웨어는 지속적으로 끊임없이 입력되는 데이터를 정확하게 실시간으로 처리하고 응용 서비스에서 질의하는 결과를 획득해서 전달하여야 한다. 이와 같은 지속적으로 입력되는 대량의 데이터 스트림을 처리하기 위해서 데이터 스트림 관리 시스템(Data Stream Management System: DSMS)을 개발하기 위한 연구가 진행되고 있다. 기존에 연구되는 데이터 스트림에 대한 알고리즘은 대부분 연속 질의 결과들 사이의 평균 오차를 줄이고, 부하 발생 시 데이터의 우선순위에 따라 버리는 것에 초점이 맞추어져 있다. 본 논문에서는 RFID EPC 라는 데이터 특성에 맞추어 다중버퍼를 이용함으로써 고속의 데이터 처리 능력을 얻고, 각 버퍼마다 일정한 규칙을 통해 질의에 있어서도 빠른 대응을 할 수 있는 알고리즘을 제안한다. 본 논문은 현재 DSMS의 관련 연구와 고속 데이터 처리의 필요성을 말하고, 제안하는 알고리즘 설명과 시뮬레이션을 통해 단일버퍼와 다중버퍼일 경우 데이터 처리 속도 성능 평가와 제안한 알고리즘에 맞도록 버퍼가 생성 되는지 테스트하는 것으로 구성된다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.342-344
• /
• 2023

클라우드 시장의 성장과 마이크로 서비스 접근식이 제기됨에 따라 IT인프라를 관리하기 위한 연구가 최근 활발히 이루어지고 있다. 하지만 고도화 및 분산된 환경에서 관찰 가능성 응용을 확보하기 어렵다는 문제점을 가지고 있다. 따라서 본 연구에서는 모니터링 시스템을 통한 데이터 분석 중 수집한 데이터의 분석이 난해하다는 문제를 해결하기 위한 방법을 제안한다. 제안된 방법은 NAB 데이터셋을 대상으로 STUMPY를 이용하여 데이터를 시각화하고, CNN을 이용하여 분류 작업을 수행한다. 분류를 수행한 데이터셋은 이상치 데이터와 이상 전조 데이터, 정상 데이터셋으로 분류하여 데이터셋을 구성한다. 구성한 학습 데이터셋에 대해 훈련을 마친 딥러닝 모델은 부하 테스트 환경에서 수집한 데이터에 대한 그래프 패턴을 분석하여 이상치 데이터와 이상 전조 데이터를 탐지한다.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.46 no.3
• /
• pp.64-76
• /
• 2009

Adaptive Random Testing through Iterative Partitioning(IP-ART) is one of Adaptive Random Testing(ART) techniques. IP-ART uses an iterative partitioning method for input domain to improve the performances of early-versions of ART that have significant drawbacks in computation time. Another version of IP-ART, named with EIP-ART(IP-ART with Enlarged Input Domain), uses virtually enlarged input domain to remove the unevenly distributed parts near the boundary of the domain. EIP-ART could mitigate non-uniform test case distribution of IP-ART and achieve relatively high performances in a variety of input domain environments. The EIP-ART algorithm, however, have the drawback of higher computation time to generate test cases mainly due to the additional workload from enlarged input domain. For this reason, a revised version of IP-ART without input domain enlargement needs to improve the distribution of test cases to remove the additional time cost. We explore three smoothing algorithms which influence the distribution of test cases, and analyze to check if any performance improvements take place by them. The simulation results show that the algorithm of a restriction area management achieves better performance than other ones.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.5
• /
• pp.3336-3344
• /
• 2014

The purpose of this study is small scale coagulation module is developed and demonstrated through a lab-scale test. Recent as a sewage treatment rate increases, have heightened the interest in the necessity on the nonpoint source and developing a small processing unit has been increased. Coagulation sedimentation module in this study is additional growth of floc through swirling in the outside zone, reduction of microstructure floc number and the internal settling zone through vertical/level flow complex sedimentation method after the coagulation process precipitation method as an effective high separation efficiency can be divided was also assessed. Coagulation sedimentation module can increase the load factor was 4.4 times compared to conventional clarifier base on the same volume and surface area through vertical/level flow. In this study, this process was selected formation and maintenance of swirling and uniform flow distribution in the internal settling zone as an important design factor, to derive its FLUENT was used to characteristics of the flow model. Through the simulation of swirling, influent velocity, dimensions of external basin, hopper depth of bottom cone was determined and through analysis of velocity distribution, flow distribution detailed specifications are derived like as diameter and number of effluent hole. Lab-scale($120{\ell}/hr$) test results, influent of 300~800 NTU to less than 10 NTU without polymer feeding was able to operate in the 20minutes retention time(surface loading rate $37.3m^3/m^2$-day), and through analysis FLUENT the possibility of using design parameters were derived.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.10
• /
• pp.86-92
• /
• 2016

A high-capacity battery installed in a hybrid vehicle or electric vehicle is used to power, or as a power supply for, electric sub-assemblies. In order to use a high-capacity battery as a power supply for electric sub-assemblies, such as an electronic control unit or for lighting, radio, and navigation, there is a need for a DC converter that changes a high voltage of 240-400V to a low voltage of 12-14V, which is done with a low-voltage DC-DC converter (LDC). An LDC undergoes long-term aging so as to reduce latent defects in the production process. With regard to the usual aging method, an LDC is a DC-DC converter. So, a DC power supply is connected and used as input, and a programmable DC electronic load is the output. For stable operation, a product having a larger capacity by 10% (compared to an LDC) is used, and has a structure where electric power is dissipated into 100% heat. So, there is a problem with volume, based on the use of two pieces of equipment to test the LDC, and another problem based on the generation of heat in the programmable DC electronic load. Hence, this paper suggests a load test method as a way of recycling, where a significant portion of the electricity dissipated as heat in a load tester is returned as input. The method realizes savings of 80% or more in the electricity dissipated as heat through improvement in the efficiency of the recycling load tester.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.6
• /
• pp.1218-1224
• /
• 2017

With the increasing risk in building liquefied natural gas carriers (LNGC), pre-simulation of various scenarios is required for system integration and safe operation. In particular, the power management system (PMS) is an important part of the LNGC; it works in tight integration with the power control systems to achieve the desired performance and safety. To verify and improve unpredicted errors, we implemented a simulation model of power generation and consumption for testing PMS based on software-in-the-loop (SIL) method. To control and verify the PMS, numeric and physical simulation modeling was undertaken utilizing MATLAB/Simulink. In addition, the simulation model was verified with a load sharing test scenario for a sea trial. This simulation allows shipbuilders to participate in new value-added markets such as commissioning, installation, operation, and maintenance.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.7
• /
• pp.1455-1462
• /
• 2012

In this paper, a BCD process based high-reliability 24-bit dual-port eFuse OTP Memory for PMICs is designed. We propose a comparison circuit at program-verify-read mode to test that the program datum is correct by using a dynamic pseudo NMOS logic circuit. The comparison result of the program datum with its read datum is outputted to PFb (pass fail bar) pin. Thus, the normal operation of the designed OTP memory can be verified easily by checking the PFb pin. Also we propose a sensing margin test circuit with a variable pull-up load out of consideration for resistance variations of programmed eFuse at program-verify-read mode. We design a 24-bit eFuse OTP memory which uses Magnachip's $0.35{\mu}m$ BCD process, and the layout size is $289.9{\mu}m{\times}163.65{\mu}m$ ($=0.0475mm^2$).

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.5
• /
• pp.153-160
• /
• 2020

Energy Consumption Predictions for Industries has a prominent role to play in the energy management and control system as dynamic and seasonal changes are occurring in energy demand and supply. This paper introduces and explores the steel industry's predictive models of energy consumption. The data used includes lagging and leading reactive power lagging and leading current variable, emission of carbon dioxide (tCO2) and load type. Four statistical models are trained and tested in the test set: (a) Linear Regression (LR), (b) Radial Kernel Support Vector Machine (SVM RBF), (c) Gradient Boosting Machine (GBM), and (d) Random Forest (RF). Root Mean Squared Error (RMSE), Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) are used for calculating regression model predictive performance. When using all the predictors, the best model RF can provide RMSE value 7.33 in the test set.