• Journal of the Korean Crystal Growth and Crystal Technology
• /
• v.33 no.6
• /
• pp.216-225
• /
• 2023

This paper describes the finite element analysis and die design change of cold heading punching process to increase the cold forging tool life and reduce the tool wear and stress concentration. Through this study, the optimization of punch tool design has been studied by an analysis of tool stress and wear distribution to improve the tool life. Plastic deformation analysis was carried out in order to understand the cold heading process between tool and workpiece stress distribution. Cold heading punch die design was set up to each process with different four types analysis progressing, the cold heading punch dies shapes with combination of point angle and punch edge corner radius shapes of cold forging dies, punch die material properties and frictional coefficient. The design parameters of point angle and corner radius of punch die geometry, die material properties and frictional coefficient were selected to apply optimization with the DoE (design of experiment) and Taguchi method. DoE and Taguchi method was performed to optimize the cold heading punch die design parameters optimization for bolt head cold forging process, it was possible to expect an reduce the cold heading punch die wear to the 37 % compared with current using cold heading punch in the shop floor.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.30 no.5D
• /
• pp.497-504
• /
• 2010

In this paper, the author presents a reliability estimation technique to analyze the effects of traffic loads on pavement mean life based on the national highway database of Suwon and Uijeongbu region from 1999 to 2008. The estimation of the mean life, its standard deviation and reliability for pavement sections are calculated by using an appropriate distribution, Lognormal distribution, based on reliability theory. Furthermore, the probability paper method and Maximum likelihood estimation are both used to estimate parameters. The author found that mean life of newly constructed sections and over-layed sections is about 6.5 to 7.9 years and 7.3 to 9.1 years, respectively. The author also ascertained that the results of cumulative failure probability for pavement life between the proposed methods and observed data are similar. Such an assessment methodology and measures based on reliability theory can provide useful information for maintenance plans in pavement management systems as long as additional life data on pavement sections are accumulated.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.18 no.2
• /
• pp.149-160
• /
• 2015

This study aims to estimate the forest volumes of the upper region of Nam-Han River in ecosystem zoning by forest types and age classes, and to suggest the optimal estimation method through the comparison of the standard errors according to the spatial unit. In the estimation of forest volumes, we used both of direct estimation, which uses sample plots of the target area only, and synthetic estimation, which includes sample plots of the expanded areas as well as those of the target area. As for the spatial expansion, we applied four standards for synthetic estimator: Mountainous zone, Neighbor ecosystem region, Gangwon province, and Buffer zone. The results show that average forest volume per ha, calculated by direct estimation, was $143.5m^3/ha$, while that by synthetic estimation with each standard, was estimated at $146.9m^3/ha$ by Gangwon province, $144.8m^3/ha$ by Buffer zone, $139.8m^3/ha$ by Neighbor ecosystem region, and $138.6m^3/ha$ by Mountainous zone, respectively. The standard errors of direct estimation was $1.79m^3/ha$, while those of synthetic estimation showed not a great difference among the errors. Meanwhile, considering the standard errors by forest type, the lowest was ${\pm}2.3m^3/ha$ of broad-leaved forest, followed by ${\pm}3.3m^3/ha$ of mixed forest, and ${\pm}4.8m^3/ha$ of coniferous forest.

• Journal of the Korean Crystal Growth and Crystal Technology
• /
• v.8 no.4
• /
• pp.592-598
• /
• 1998

AZO(Aluminum doped Zinc Oxide) thin films were fabricated by reactive DC magnetron sputtering method using zinc metal target (Al 2%) and zinc oxide target ($Al_2O_3\;2%$) respectively. The intermediate condition with optimum transmittance and conductivity was obtained by controlling the sputtering parameters. Oxygen gas ratio for this condition was $0.5{\times}10^{-2}~1.0{\times}10^{-2}$ in oxide target and. In case of metal target, this optimum oxygen gas ratio at the applied power of 0.6 kW and 1.0 kW was 0.215~0.227 and 0.305~0.315, respectively. The resistivity of AZO film deposited was obtained $1.2~1.4{\times}10^{-3} {\Omega}{\cdot}$cm as deposited state regardless of target species.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.26 no.1
• /
• pp.104-112
• /
• 1989

A design automation method for sequential circuits implementation by mans of PLA is discussed, and an optimal state-code assignment algorithm to minimize the PLA area is proposed. In order to design sequential circuit automatically, DASL (Design Automation Support Language) [8] which is easy to describe and powerful to synthesize, is proposed and used to describe sequential circuit, An optimal statecode assignment algorithm which considers next states and outputs simultaneously is proposed, and by adopting this algorithm to various examples, the area of PLA is reduced by 10% comparing privious methods. This system is constructed to design microinstruction, FSM, VLSI control part synthesis.

• Water for future
• /
• v.9 no.1
• /
• pp.101-105
• /
• 1976

This Study is developed in order to determine the probability of a raintall depth of short duration in Seoul as considering the profect life and the factor of safety of hydraulic structures. The raw annual maximum rainfall data are selected from 1915 to 1974 about short duration (10min-120min.) in Seoul. The selected data are treated by frequency analysis, and the hypothesis that the distribution fuction of the raw data is normal Distribution is performed by chi-square test that signifcance level has 5%. With the parameters (mean and standard deviation) of the accepted distribution function, the probabilitn of a rainfall depth can be easily determined on the graph which is made on this paper.

• Water for future
• /
• v.24 no.1
• /
• pp.63-71
• /
• 1991

This study represents the methodology for feasibility analysis of small hydro power plants. Cumulative density function of Weibull distribution and Thi-essen method were adopted to beside flow duration curve at candidate sites. The performance prediction model and construction cost estimation model for tunnel-type small hydro power plants were developed. Eight candidate sites existing on Han river selected and surveyed for actual sites reconnaissance. The performance characteristics and economical feasibility for these sites were analyzed by using developed models. As a result, it was found that the optimum design flowrates with the lowest unit generation cost for tunnel-type small hydro power plants were the flowrate concerning with between 20 % and 30 % of time ratio on the flow duration curve. Additionally, primary design specifications such as design flowrate, effective head, capacity, annual average load factor, annual electricity production were estimated and discussed for surveyed sites.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.229-229
• /
• 2023

강화학습은 에이전트(agent)가 주어진 환경(environment)과의 상호작용을 통해서 상태(state)를 변화시켜가며 최대의 보상(reward)을 얻을 수 있도록 최적의 행동(action)을 학습하는 기계학습법을 의미한다. 최근 알파고와 같은 게임뿐만 아니라 자율주행 자동차, 로봇 제어 등 다양한 분야에서 널리 사용되고 있다. 상수도관망 분야의 경우에도 펌프 운영, 밸브 운영, 센서 최적 위치 선정 등 여러 문제에 적용되었으나, 설계에 강화학습을 적용한 연구는 없었다. 설계의 경우, 관망의 크기가 커짐에 따라 알고리즘의 탐색 공간의 크기가 증가하여 기존의 최적화 알고리즘을 이용하는 것에는 한계가 존재한다. 따라서 본 연구는 강화학습을 이용하여 상수도관망의 구성요소와 환경요인 간의 복잡한 상호작용을 고려하는 설계 방법론을 제안한다. 모델의 에이전트를 딥 강화학습(Deep Reinforcement Learning)으로 구성하여, 상태 및 행동 공간이 커 발생하는 고차원성 문제를 해결하였다. 또한, 해당 모델의 상태 및 보상으로 절점에서의 압력 및 수요량과 설계비용을 고려하여 적절한 수량과 수압의 용수 공급이 가능한 경제적인 관망을 설계하도록 하였다. 모델의 행동은 실제로 공학자가 설계하듯이 절점마다 하나씩 차례대로 다른 절점과의 연결 여부를 결정하는 것으로, 이를 통해 관망의 레이아웃(layout)과 관경을 결정한다. 본 연구에서 제안한 방법론을 규모가 큰 그리드 네트워크에 적용하여 모델을 검증하였으며, 고려해야 할 변수의 개수가 많음에도 불구하고 목적에 부합하는 관망을 설계할 수 있었다. 모델 학습과정 동안 에피소드의 평균 길이와 보상의 크기 등의 변화를 비교하여, 제안한 모델의 학습 능력을 평가 및 보완하였다. 향후 강화학습 모델을 통해 신뢰성(reliability) 또는 탄력성(resilience)과 같은 시스템의 성능까지 고려한 설계가 가능할 것으로 기대한다.

• Analytical Science and Technology
• /
• v.7 no.1
• /
• pp.25-31
• /
• 1994

As for analytes which did not represent the differences of the selectivity on the stationary and mobile phase, secondary chemical equilibrium theory was applied to study pH effects on the separation of alkylphenols. Mobile phase was consisted of an aqueous sodium carbonate-bicarbonate buffer and acetonitrile. The maximum selectivity for adjacent peak pairs was predicted from those values of $k^{\prime}_{HA}/k^{\prime}_{A-}$ and ${\Delta}pK$. The optimum pH determined by this method was 11.18 pH and solvent selectivity were considered at the same time to invoke the full range of selectivity effects possible for separations. Quaternary mixture composed of methanol, acetonitrile, tetrahydrofuran and water was adjusted to optimum pH 11.18. As the statistical simplex technique of an overlapping resolution map (ORM) was used to predict the optimized solvent system. The optimum solvent, which gives complete separation of alkylphenols, was determined as follws MeOH : ACN : THF = 14.4 : 81.8 : 2.8.

• The KIPS Transactions:PartD
• /
• v.8D no.2
• /
• pp.175-180
• /
• 2001

This paper presents the optimal fraction of validation set to obtain a prediction accuracy of software failure count or failure time in the future by a neuro-fuzzy system. Given a fixed amount of training data, the most popular effective approach to avoiding underfitting and overfitting is early stopping, and hence getting optimal generalization. But there is unresolved practical issues : How many data do you assign to the training and validation set\ulcorner Rules of thumb abound, the solution is acquired by trial-and-error and we spend long time in this method. For the sake of optimal fraction of validation set, the variant specific fraction for the validation set be provided. It shows that minimal fraction of the validation data set is sufficient to achieve good next-step prediction. This result can be considered as a practical guideline in a prediction of software reliability by neuro-fuzzy system.