• 한국정보통신학회논문지
• /
• 제7권2호
• /
• pp.284-292
• /
• 2003

본 논문에서는 LMS 알고리즘을 이용하여 적응 시스템을 제어하는 경우, 최적 매개변수의 선택 폭을 늘리기 위한 초기치의 설정방법의 수치적 검토를 행했다. 초기치 설정은 대략의 값을 임의적으로 선택하는 일반적인 방법과 직접법에 의해 얻어진 근사적인 해를 초기치로서 가하는 방법을 이용하였으며, 이들을 최적지향성합성 문제에 적용하여 초기치가 매개변수의 선택 폭, 자승평균오차의 수렴속도, 그리고 수렴과정에서의 안정성 등에 미치는 영향을 비교하였다. 수치실험결과, 직접법을 이용한 초기치 선택방법은 일반적인 선택방법에 비해 매개변수의 선택 폭이 넓어짐은 물론 수렴성, 안정성 그리고 오차 개선 능력도 탁월함을 나타내었다.

• 한국물환경학회지
• /
• 제25권6호
• /
• pp.979-983
• /
• 2009

The mean pH of wastewater discharged from the plating process is 2, so a less amount of alkali is required to raise pH 2 to 5. In addition, if sodium sulfite is used to raise pH 5 to 9 in the secondary treatment, caustic soda or slaked lime is not necessary or only a small amount is necessary because sodium sulfite is alkali. Thus, it is considered desirable to use only $FeSO_4{\cdot}7H_2O$ in the primary treatment. At that time, the free cyanide removal rate was highest as around 99.3%, and among heavy metals, Ni showed the highest removal rate as around 92%, but zinc and chrome showed a low removal rate. In addition, the optimal amount of $FeSO_4{\cdot}7H_2O$ was 0.3g/L, at which the cyanide removal rate was highest. Besides, the free cyanide removal rate was highest when pH value was 5. Of cyanide removed in the primary treatment, the largest part was removed through the precipitation of ferric ferrocyanide: $[Fe_4(Fe(CN)_6]_3$, and the rest was precipitated and removed through the production of $Cu_2[Fe(CN)_6]$, $Ni_2[Fe(CN)_6]$, CuCN, etc. Furthermore, it appeared more effective in removing residual cyanide in wastewater to mix $Na_2SO_3$ and $Na_2S_2O_5$ at an optimal ratio and put the mixture than to put them separately, and the optimal weight ratio of $Na_2SO_3$ to $Na_2S_2O_5$ was 1:2, at which the oxidative decomposition of residual cyanide was the most active. However, further research is required on the simultaneous removal of heavy metals such as chrome and zinc.

• 반도체디스플레이기술학회지
• /
• 제20권2호
• /
• pp.92-97
• /
• 2021

The importance of processes in cryogenic environments is increasing in a way to address problems such as critical dimension (CD) narrow and bottlenecks in micro-processing. Accordingly, in this paper, we proceed with the design and analysis of Electrostatic Chuck(ESC) and Coolant in cryogenic environments, and present optimal model conditions to provide the temperature distribution analysis of ESC in these environments and the appropriate optimal design. The wafer temperature uniformity was selected as the reference model that the operating conditions of the refrigerant of the liquid nitrogen in the doubled aluminum path were excellent. Design of simulation (DOS) was carried out based on the wheel settings within the selected reference model and the classification of three mass flow and diameter case, respectively. The comparison between factors with p-value less than 0.05 indicates that the optimal design point is when five turns of coolant have a flow rate of 0.3 kg/s and a diameter of 12 mm. ANOVA determines the interactions between the above factor, indicating that mass flow is the most significant among the parameters of interests. In variable selection procedure, Case 2 was also determined to be superior through the two-Sample T-Test of the mean and variance values by dividing five coolant wheels into two (Case 1 : 2+3, Case 2: 3+2). Finally, heat transfer analysis processes such as final difference method (FDM) and heat transfer were also performed to demonstrate the feasibility and adequacy of the analysis process.

• International Journal of Quality Innovation
• /
• 제9권3호
• /
• pp.71-91
• /
• 2008

The definition of cycle time is the time from the wafer start to the wafer output. It usually takes one or two months to get the product since customer decides to produce it. The cycle time is a critical factor for customer satisfaction because it represents the response time to the market. Long cycle time reflects the ineffective investment for the capital. The cycle time is very important for foundry because long cycle time will cause customer unsatisfied and the order loss. Consequently, all of the foundries put lots of human source in the cycle time improvement. Usually, we make decisions based on the experience in the cycle time management. We have no mechanism or theory for cycle time management. We do work-in-process (WIP) management based on turn rate and standard WIP (STD WIP) set by experiences. But the experience didn't mean the optimal solution, when the situation changed, the cycle time or the standard WIP will also be changed. The experience will not always be applicable. If we only have the experience and no mechanism, management will not be work out. After interview several foundry fab managers, all of the fab can't reflect the situation. That is, all of them will have an impact period after product mix or utilization varied. In this study, we want to develop a formula for standard WIP and use statistical process control (SPC) concept to set WIP upper/lower limit level. When WIP exceed the limit level, it will trigger action plans to compensate WIP Profile. If WIP Profile balances, we don't need too much WIP. So WIP level could be reduced and cycle time also could be reduced.

• 한국전자통신학회논문지
• /
• 제9권12호
• /
• pp.1337-1344
• /
• 2014

디지털 가입자회선과 같은 통신에서 반향 제거기의 목적은 수신 경로에서 하이브리드 회로에 의해 누출되는 전송 신호를 보상하는 것이다. 일반적으로 전이중 통신에서 사용되는 반향 제거기는 지엽적인 신호에 의해 동작되는 적응 시스템이며, 최소 평균 제곱 알고리듬으로 구현된 반향 제거기가 많이 사용되어 왔지만 적은 계산 양의 장점을 가지는 반면에 느린 수렴 성능을 보인다. 또한, 반향 제거기의 길이는 성능과 수렴속도에 직접적인 영향을 미치며, 긴 시간동안 변화하는 반향을 제거하기 위해서는 반향 제거기의 계수 개수가 커야 하는데, 이것은 적응 필터의 수렴 속도를 감소시킨다. 본 논문에서는 통신 채널에서의 반향 제거에 대한 새로운 방법을 제안한다. 제안한 방법은 순방향 알고리듬과 역방향 알고리듬의 가중 결합으로 구성된 양방향 최소 평균 제곱 알고리듬을 사용하여 반향 제거기의 최적의 계수를 계산한다. 마지막으로 수학적 해석 및 모의실험을 통해 제안한 반향 제거기가 계산 양의 증가가 없이 거의 동일한 계산 양으로 기존의 반향 제거기보다 수렴 속도가 빠르다는 사실을 확인하였다.

• 한국의학물리학회:학술대회논문집
• /
• 한국의학물리학회 2002년도 Proceedings
• /
• pp.113-115
• /
• 2002

The Monte Carlo simulation method is a numerical solution to a problem that models objects interacting with other objects or their environment based upon simple object-object or object-environment relationships. In spite of its great accuracy, It was turned away because of long calculation time to simulate a model. But, it is used to simulate a linear accelerator frequently with the advance of computer technology. To simulate linear accelerator in Monte Carlo simulations, there are many parameters needed to input to Monte Carlo code. These data can be supported by a linear accelerator manufacturer. Although the model of a linear accelerator is the same, a different characteristic property can be found. Thus, we performed a commissioning process of 6MV photon beam in Varian 2300C/D model with BEAM/EGS4 Monte Carlo code. The head geometry data were put into BEAM/EGS4 data. The mean energy and energy spread of the electron beam incident on the target were varied to match Monte Carlo simulations to measurements. TLDs (thermoluminescent dosimeter) and radiochromic films were employed to measure the absorbed dose in a water phantom. Beam profile was obtained in 40cm${\times}$40cm field size and Depth dose was in 10cm${\times}$10cm. At first, we compared the depth dose between measurements and Monte Carlo simulations varying the mean energy of an incident electron beam. Then, we compared the beam profile with adjusting the beam radius of the incident electron beam in Monte Carlo simulation. The results were found that the optimal mean energy was 6MV and beam radius of 0.1mm was well matched to measurements.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권7호
• /
• pp.3370-3392
• /
• 2017

Next Generation Beyond 4G/5G systems will rely on the deployment of small cells over conventional macrocells for achieving high spectral efficiency and improved coverage performance, especially for indoor and hotspot environments. In such heterogeneous networks, the expected performance gains can only be derived with the use of efficient interference coordination schemes, such as Fractional Frequency Reuse (FFR), which is very attractive for its simplicity and effectiveness. In this work, femtocells are deployed according to a spatial Poisson Point Process (PPP) over hexagonally shaped, 6-sector macro base stations (MeNBs) in an uncoordinated manner, operating in hybrid mode. A newly introduced intermediary region prevents cross-tier, cross-boundary interference and improves user equipment (UE) performance at the boundary of cell center and cell edge. With tools of stochastic geometry, an analytical framework for the signal-to-interference-plus-noise-ratio (SINR) distribution is developed to evaluate the performance of all UEs in different spatial locations, with consideration to both co-tier and cross-tier interference. Using the SINR distribution framework, average network throughput per tier is derived together with a newly proposed harmonic mean, which ensures fairness in resource allocation amongst all UEs. Finally, the FFR network parameters are optimized for maximizing average network throughput, and the harmonic mean using a fair resource assignment constraint. Numerical results verify the proposed analytical framework, and provide insights into design trade-offs between maximizing throughput and user fairness by appropriately adjusting the spatial partitioning thresholds, the spectrum allocation factor, and the femtocell density.

• 대한기계학회논문집A
• /
• 제29권9호
• /
• pp.1217-1224
• /
• 2005

To improve the accuracy of a metamodel, additional sample points can be selected by using a specified criterion, which is often called sequential sampling approach. Sequential sampling approach requires small computational cost compared to one-stage optimal sampling. It is also capable of monitoring the process of metamodeling by means of identifying an important design region for approximation and further refining the fidelity in the region. However, the existing critertia such as mean squared error, entropy and maximin distance essentially depend on the distance between previous selected sample points. Therefore, although sufficient sample points are selected, these sequential sampling strategies cannot guarantee the accuracy of metamodel in the nearby optimum points. This is because criteria of the existing sequential sampling approaches are inefficient to approximate extremum and inflection points of original model. In this research, new sequential sampling approach using the sensitivity of metamodel is proposed to reflect the response. Various functions that can represent a variety of features of engineering problems are used to validate the sensitivity approach. In addition to both root mean squared error and maximum error, the error of metamodel at optimum points is tested to access the superiority of the proposed approach. That is, optimum solutions to minimization of metamodel obtained from the proposed approach are compared with those of true functions. For comparison, both mean squared error approach and maximin distance approach are also examined.

• 멤브레인
• /
• 제31권1호
• /
• pp.52-60
• /
• 2021

SPG (Shirasu porous glass) 원통형 막을 회전 시키는 회전 막유화를 사용하여 칼슘 알지네이트 미소 구체를 제조할 때, 단분산 미소 구체를 제조하기 위한 회전 막유화 공정변수들의 최적 조건을 결정하였다. 회전 막유화의 공정 변수로는 막의 회전 속도, 막간 압력차, 연속상에 대한 분산상의 비율, 알지네이트 농도, 유화제의 농도, 안정제 농도, 가교제 농도 및 막의 세공 크기를 설정하고, 이들 변수로 제조된 알지네이트 미소 구체의 크기와 단분산성에 미치는 영향을 검토하였다. 이 결과 회전 막유화의 공정 변수들 중에서 막모듈의 회전 속도, 유화제의 농도, 가교제의 농도가 증가 할수록 미소 구체의 크기가 감소하였으며, 반면에 연속상에 대한 분산상의 비율, 막간 압력차, 알지네이트 농도가 증가할수록 미소 구체의 크기가 증가하였다. 세공 크기 3.2 ㎛인 SPG막을 사용한 회전 막유화에서 공정변수 조절을 통해 최종적으로 입자 크기가 4.5 ㎛의 단분산 알지네이트 미소 구체의 제조가 가능하였다.

• Korean Chemical Engineering Research
• /
• 제61권3호
• /
• pp.388-393
• /
• 2023

지도 학습 기반의 신경 망을 활용한 공학적 자료의 분석은 화학공학 공정 최적화, 미세 먼지 농도 추정, 열역학적 상평형 예측, 이동 현상 계의 물성 예측 등 다양한 분야에서 활용되고 있다. 신경 망의 지도 학습은 학습 자료를 요구하며, 주어진 학습 자료의 구성에 따라 학습 성능이 영향을 받는다. 빈번히 관찰되는 공학적 자료 중에는 DNA의 길이, 분석 물질의 농도 등과 같이 로그 간격으로 주어지는 자료들이 존재한다. 본 연구에서는 넓은 범위에 분포된 로그 간격의 학습 자료를 기계 학습으로 처리하는 경우, 사용 가능한 손실 함수들의 학습 성능을 정량적으로 평가하였으며, 적합한 학습 자료 구성 방식을 연구하였다. 이를 수행하고자, 100×100의 가상 이미지를 활용하여 기계 학습의 회귀 과업을 구성하였다. 4개의 손실 함수들에 대하여 (i) 오차 행렬, (ii) 최대 상대 오차, (iii) 평균 상대 오차로 정량적 평가하여, mape 혹은 msle가 본 연구에서 다룬 과업에 대해 최적의 손실 함수가 됨을 알아내었다. 또한, 학습 자료의 값이 넓은 범위에 걸쳐 분포하는 경우, 학습 자료의 구성을 로그 간격 등을 고려하여 균등 선별하는 방식이 높은 학습 성능을 보임을 밝혀내었다. 본 연구에서 다룬 회귀 과업은 DNA의 길이 예측, 생체 유래 분자 분석, 콜로이드 용액의 농도 추정 등의 공학적 과업에 적용 가능하며, 본 결과를 활용하여 기계 학습의 성능과 학습 효율의 증대를 기대할 수 있을 것이다.