• 한국통신학회논문지
• /
• 제24권8A호
• /
• pp.1187-1194
• /
• 1999

일반적으로 무선 이동 채널은 협대역 스펙트럼을 갖는 복소 랜덤 프로세스로 모델링된다. 본 논문에서는 DSP 단일칩을 사용한 페이딩 신호의 실시간 발생에 대해 기술한다. 실시간 시뮬레이터는 단말기 이동 속도, 반송파 주파수, line-of-sight 성분과 다중 경로 성분의 비, 수신 전력의 분산과 같은 시뮬레이션 파라미터를 윈도우상에서 선택할 수 있도록 설계되었다. 최소화된 DSP 연산량으로 이상적인 페이딩 신호를 발생하기 위한 알고리즘과 필터 설계시의 trade-off가 고찰되었다. 실험으로 측정된 페이딩 채널 프로세스의 통계 특성은 이론치와 거의 일치함을 확인할 수 있었다.

• Journal of the Optical Society of Korea
• /
• 제13권3호
• /
• pp.361-366
• /
• 2009

We propose the performance enhancing method optimization of an asynchronous 2.5 Gbps/1.25 Gbps optical subscriber network with inverse RZ (Return to Zero) coded downstream and NRZ (Non Return to Zero) upstream re-modulation by adjusting threshold level control of a receiver. We theoretically analyze the BER (Bit Error Rate) performance by modeling the occurrence of BER by simulation with MATLAB according to the types of downstream data. The results have shown that the normalized threshold level in an optical receiver could be saturated at 1/3 as the SNR (Signal to Noise Ratio) increases. The needed SNR for obtaining the BER $10^{-9}$ can be reduced by $\sim$5 dB by optimizing the normalized threshold level at 1/3 instead of by using the conventional receiver with threshold level of 0.5. The proposed system can be a useful technology for asynchronous optical access networks with asymmetric upstream and downstream data rates, because the improved minimum receiving power could replace a light source with a source with lower power and lower cost in an OLT (Optical Line Termination).

• Nuclear Engineering and Technology
• /
• 제56권3호
• /
• pp.980-992
• /
• 2024

Steam line break accident (SLB) in the nuclear reactor is one of the representative Non-LOCA accidents in which thermal-hydraulics and neutron kinetics are strongly coupled each other. Thus, the multi-scale and multi-physics approach is applied in this study in order to examine a realistic safety margin. An entire reactor coolant system is modelled by system scale node, whereas sub-channel scale resolution is applied for the region of interest such as the reactor core. Fuel performance code is extended to consider full core pin-wise fuel behaviour. The MARU platform is developed for easy integration of the codes to be coupled. An initial stage of the steam line break accident is simulated on the MARU platform. As cold coolant is injected from the cold leg into the reactor pressure vessel, the power increases due to the moderator feedback. Three-dimensional coolant and fuel behaviour are qualitatively visualized for easy comprehension. Moreover, quantitative investigation is added by focusing on the enhancement of safety margin by means of comparing the minimum departure from nucleate boiling ratio (MDNBR). Three factors contributing to the increase of the MDNBR are proposed: Various geometric parameters, realistic power distribution by neutron kinetics code, Radial coolant mixing including sub-channel physics model.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 하계학술대회 논문집 C
• /
• pp.926-928
• /
• 1998

A fast optimization algorithm has been evolved from a simple two stage optimal power flow(OPF) algorithm for constrained power economic dispatch. In the proposed algorithm, we consider various constraints such as power balance, generation capacity, transmission line capacity, transmission losses, security equality, and security inequality constraints. The proposed algorithm consists of four stages. At the first stage, we solve the aggregated problem that is the crude classical economic dispatch problem without considering transmission losses. An initial solution is obtained by the aggregation concept in which the solution satisfies the power balance equations and generation capacity constraints. Then, after load flow analysis, the transmission losses of an initial generation setting are matched by the slack bus generator that produces power with the cheapest cost. At the second stage we consider transmission losses. Formulation of the second stage becomes classical economic dispatch problem involving the transmission losses, which are distributed to all generators. Once a feasible solution is obtained from the second stage, transmission capacity and other violations are checked and corrected locally and quickly at the third stage. The fourth stage fine tunes the solution of the third stage to reach a real minimum. The proposed approach speeds up the coupled LP based OPF method to an average gain of 53.13 for IEEE 30, 57, and 118 bus systems and EPRI Scenario systems A through D testings.

• 한국정밀공학회지
• /
• 제17권11호
• /
• pp.191-198
• /
• 2000

Laser direct patterning of the coated photoresist (PMER-NSG31B) layer was studied to make halftone dots on gravure printing roll. The selective laser hardening of photoresist by Ar-ion laser(wavelength: 333.6~363.8nm) was controlled by the A/O modulator. The coating thickness in the range of 5~11$\mu m$ could be obtained by using the up-down directional moving device along the vertically located roll. The width, thickness and hardness of the hardened lines formed under the laser power of 200~260㎽ and irradiation time of 4.4~6.6 $\mu$sec/point were investigated after developing. The hardened width increased as the coating thickness increased. Though the hardened thickness was changed due to the effect of the developing solution, the hardened layer showed good resistance to the scratching of 2H pencil. Also, the hardened minimum line widths of 10$\mu m$ could be obtained. The change of line width was also found after etching, and the minimum line widths of 6$\mu m$ could be obtained. The hardened lines showed the good resistance to the etching solution. Finally, the experimental data could be applied to make gravure halftone dots using the developed imaging process, successfully.

• 대한전자공학회논문지SD
• /
• 제41권4호
• /
• pp.1-8
• /
• 2004

고속 저전력 디지털 시스템을 위해 클록 스큐를 최소화하고 동적 파워 소모를 줄이는 새로운 클록 분배 방법을 제안하였다. 제안된 방법은 접힌 라인구조(FCL)과 위상 섞임 회로(phase blending circuit)을 이용하여 Zero-skew 특성을 갖는다. FCL에 적합한 라인 구조를 분석하기 위해, 마이크로 스트립과 코플라너 라인을 FCL형 클록 라인으로 분배되었다. 시뮬레이션 결과는 l0㎜ 떨어져 있는 두 리시버 사이의 최대 클록 스큐가 1㎓에서 10psec보다 적고 20㎜ 떨어져 있는 두 리시버 사이의 최대 클록 스큐는 1㎓에서 60 psec보다 작음을 보였다. 또한, 공정, 전압, 온도 변화에 무관하게 클록 신호들의 스큐가 변하지 않음을 알 수 있었다.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2001년도 춘계학술대회논문집D
• /
• pp.915-920
• /
• 2001

A spent fuel problem has prevented the nuclear power from claiming to be a completely clean energy source. The nuclear transmutation technology to incinerate the long lived radioactive nuclides and produce energy during the incineration process is believed to be one or the best solutions. HYPER(Hybrid Power Extraction Reactor) is the accelerator driven transmutation system which is being developed by KAERI(Korea Atomic Energy Research Institute). Some major feature of HYPER have been developed and employed. On-power fueling concepts are employed to keep system power constant with minimum variation of accelerator power. A hollow cylinder-type metal fuel is designed for the on-line refueling concept. Lead-bismuth(Pb-Bi) is adopted as a coolant and Spallation target material. HYPER is a subcritical reactor which needs an external neutron source. 1GeV proton beam is irradiated to Lead-bismuth(Pb-Bi) target inside HYPER, and spallation neutrons are produced. When proton beams are irradiated, much heat is also deposited in the Pb-Bi target and beam window which separates Pb-Bi and accelerator vacuum. Therfore, an effective cooling is needed for HYPER target. In this paper, we performed the thermal-hydraulic analysis of HYPER target using FLUENT code, and also calculated thermal and mechanical stress of the beam window using ANSYS code.

• 조명전기설비학회논문지
• /
• 제28권2호
• /
• pp.42-47
• /
• 2014

The urban railway station is a structure that consumes a large amount of electric energy. Thus, the rationalization of using electric power is acutely demanded, but statistical data and design criteria that consider operating power load characteristics and structures, are very insufficiency. Therefore, this study investigated and analyzed that electricity consumption a year, characteristics, gross floor area of the station, structure, and etc, on the basis of Seoul metro station 1~4 line. Through regression analysis theory, we verified the overall features and the main tendency by analysis of specific parameter value(average, maximum, minimum, etc). The object of this study is the analysis about power density considering structure of the urban railway station, managing electric energy for the rationalization and setting a new standard of maintenance and construction.

• Journal of Power Electronics
• /
• 제9권6호
• /
• pp.823-834
• /
• 2009

This paper proposes a high-efficiency single-stage ac/dc converter. The proposed converter features low voltage stresses and reduced switching losses. It operates at the boundary of discontinuous- and continuous-conduction modes by employing variable switching frequency control. The turn-on switching loss of the switch can be reduced by turning it on when the voltage across it is at a minimum. The voltage across the bulk capacitor is independent of the output loads and maintained within the practical range for the universal line input, so the problem of high voltage stress across the bulk capacitor is alleviated. Moreover, the voltage stress of the output diodes is clamped to the output voltage, and the output diodes are turned off at zero-current. Thus, the reverse-recovery related losses of the output diodes are eliminated. The operational principles and circuit analysis are presented. A prototype circuit was built and tested for a 150 W (50V/3A) output power. The experimental results verify the performance of the proposed converter.

• 전자공학회논문지A
• /
• 제32A권1호
• /
• pp.70-78
• /
• 1995

In order to implement broadband RAM's(Radar Absorbing Materials) made up of multiple dielectricl layers, the design variables such as the dielectrci constaints, the depths and the loss tangents of dielectric are optimized. The wave impedances regarding the reflective wave are found in dielectrics, input impedances and reflection coefficients with multiple dielectric layers are derived from the transmission line circuit theory. Finally, minimum average reflective power and optimum design variables are obtained by applying the numerical technique, called modified Powell method. In case of four dielectric layers with inequality constraints in design variables, a quite favourable and feasible result with the total depth of 1.1 cm, the average reflective power of 0.85% over the bradband frequency range is obtained for a specific example.