• 상하수도학회지
• /
• 제35권1호
• /
• pp.93-100
• /
• 2021

Forward osmosis (FO) process is a chemical potential driven process, where highly concentrated draw solution (DS) is used to take water through semi-permeable membrane from feed solution (FS) with lower concentration. Recently, commercial FO membrane modules have been developed so that full-scale FO process can be applied to seawater desalination or water reuse. In order to design a real-scale FO plant, the performance prediction of FO membrane modules installed in the plant is essential. Especially, the flux prediction is the most important task because the amount of diluted draw solution and concentrate solution flowing out of FO modules can be expected from the flux. Through a previous study, a theoretical based FO module model to predict flux was developed. However it needs an intensive numerical calculation work and a fitting process to reflect a complex module geometry. The idea of this work is to introduce deep learning to predict flux of FO membrane modules using 116 experimental data set, which include six input variables (flow rate, pressure, and ion concentration of DS and FS) and one output variable (flux). The procedure of optimizing a deep learning model to minimize prediction error and overfitting problem was developed and tested. The optimized deep learning model (error of 3.87%) was found to predict flux better than the theoretical based FO module model (error of 10.13%) in the data set which were not used in machine learning.

• 한국정밀공학회지
• /
• 제24권6호
• /
• pp.131-137
• /
• 2007

In metal casting process, it is very difficult to predict the form accuracy of cast part and reduce repeatability error. In this study, the variations of form accuracy were measured in the process of metal cast prototyping, where RP part is manufactured from CAD model in the first, and then, wax part is cast in the vacuum environment using the RP part as master model, and finally metal prototype is cast using ceramic mold and the wax part as pattern. To investigate the variations of form accuracy, the averages and standard deviations of error distribution of the parts measured by 3D scanner were compared. It was observed that the biggest shrinkage is generated during the extraction of wax part in the second step and the biggest deterioration of form accuracy is generated during the metal part casting in the last step.

• 한국건설관리학회논문집
• /
• 제18권6호
• /
• pp.98-108
• /
• 2017

최근 모듈러 공법을 활용한 중 고층 건축 기술이 개발되면서 현장 시공 오차 관리의 중요성이 부각되었다. 모듈러 공법은 공장 생산 특성 때문에 현장에서 시공 오차를 조정하는 것이 매우 제한적이다. 따라서 시공 오차를 사전에 예방하기 위해서는 모듈러 공법 특성을 반영한 관리 방안이 필요하며 각 단계별 공사 참여자의 의사 결정이 중요하다. 이에 본 연구에서는 모듈러 공사 시공 오차 요인을 분석하고, 시공 프로세스 기반의 시공 오차 관리 방안 의사 결정 지원 모델을 제안하였다. 본 연구의 결과물은 모듈러 공사 참여자가 시공 오차 관리 방안을 도출하는데 가이드라인 역할을 할 수 있을 것이라 기대한다.

• 품질경영학회지
• /
• 제22권2호
• /
• pp.41-50
• /
• 1994

Consider a canning process where cans are filled with an expensive ingredient. Cans weighting above the specified limit are sold in a regular market for a fixed price, and underfilled cans are emptied and refilled at the expense of a reprocessing cost. In this paper, the effect of measurement error on the determination of the optimal process mean for a canning process is examined. It is assumed that the quantity X of ingredient in a can is normally distributed with unknown mean and known variance, and the observed value Y of X is also normally distributed with known mean and variance. A profit model is constructed which involves selling price. cost of ingredients, reprocessing cost. and cost from an accepted nonconforming can, and methods of finding the optimal process mean and the cutoff value on Y are presented. It is shown that the optimal process mean increases. and the expected profit decreases when the measurement error is relatively large in comparison to the process variance.

• Bulletin of the Korean Chemical Society
• /
• 제22권3호
• /
• pp.251-252
• /
• 2001

• 한국음향학회지
• /
• 제22권8호
• /
• pp.680-686
• /
• 2003

디지털 도파관 모델은 악기의 물리적 모델링에 사용되는 일반적인 방법이다. 디지털 도파관 모델에서 파동의 움직임은 시간 또는 공간을 기준으로 해석 가능하다. 음의 샘플링이 시간을 기준으로 이루어지므로 악기 모델은 시간에 의한 파동의 움직임으로 묘사되는 것이 일반적이다. 본 논문에서는 현에 대한 공간 기준의 디지털 도파관 모델에 악기 몸체 모델을 추가해 악기 음을 합성하였다. 그렇게 함으로써 합성 음의 음질을 향상시키고 악기 모델의 음색 조절 변수들을 효과적으로 처리할 수 있었다. 공간 기준 샘플링에서 현 및 몸체에서 발생하는 미소 지연 오차에 대해 설명하고 FD (Fractional Delay) 필터를 이용해 미소 지연을 처리하는 방법을 보였다. 그리고 지연에 수에 따른 합성음의 변화를 설명하고 그 결과를 시간 기준 디지털 도파관 모델과 비교하였다.

• 한국전기전자재료학회논문지
• /
• 제23권11호
• /
• pp.831-836
• /
• 2010

In this paper, Plasma Enhanced Chemical Vapor Deposition (PECVD) $SiO_2$ film properties are modeled using statistical analysis and neural networks. For systemic analysis, Box-Behnken's 3 factor design of experiments (DOE) with response surface method are used. For characterization, deposited film thickness and film stress are considered as film properties and three process input factors including plasma RF power, flow rate of $N_2O$ gas, and flow rate of 5% $SiH_4$ gas contained at $N_2$ gas are considered for modeling. For film thickness characterization, regression based model showed only 0.71% of root mean squared (RMS) error. Also, for film stress model case, both regression model and neural prediction model showed acceptable RMS error. For sensitivity analysis, compare to conventional fixed mid point based analysis, proposed sensitivity analysis for entire range of interest support more process information to optimize process recipes to satisfy specific film characteristic requirements.

• 한국기계가공학회지
• /
• 제12권5호
• /
• pp.9-15
• /
• 2013

Roll-to-roll printing technology has lately become a subject of special interests in the field of printed electronics. Since this technology has the advantage that continuous and mass production is possible. And for high precision register control is required in multi-layer printing to produce the electronic devices, this is one of the most important technologies in roll-to-roll printing technology. Register error could be generated by various reasons like eccentricity of roll and thermal deformation due to temperature variation in drying section. In this study, the effect of tension variation on the register was analyzed. The results of these analyses show that it is essential to consider the tension disturbance which is generated by the change of temperature in drying section, and conventional register model has limitation to estimate the register error. In order to overcome the limitation of the register model, advanced register model based on the SI process was developed. Also, the performance of the developed model was verified experimentally.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2013년도 춘계학술대회 논문집
• /
• pp.454-459
• /
• 2013

For the purpose of rapid development and superior design quality assurance, sophisticated finite element model for SOM(Spatial Optical Modulator) piezo actuator of MOEMS device has been developed and evaluated for the accuracy of dynamics and residual stress analysis. Parametric finite element model is constructed using ANSYS APDL language to increase the design and analysis performance. Geometric dimensions, mechanical material properties for each thin film layer are input parameters of FE model and residual stresses in all thin film layers are simulated by thermal expansion method with psedu process temperature. $6^{th}$ mask design samples are manufactured and $1^{st}$ natural frequency and 10V PZT driving displacement are measured with LDV. The results of experiment are compared with those of the simulation and validate the good agreement in $1^{st}$ natural frequency within 5% error. But large error over 30% occurred in 10V PZT driving displacement because of insufficient PZT constant $d_{31}$ measurement technology.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 1998년도 가을 학술발표논문집 Vol.25 No.2 (2)
• /
• pp.327-329
• /
• 1998

Up-Propagation is an algorithm for inverting and learning neural network generative models. Sensory input is processed by inverting a model that generates patterns from hidden variables using top-down connections. The inversion process is iterative, utilizing a negative feedback loop that depends on an error signal propagated by bottom-up connections. The error signal is also used to learn the generative model from examples. the algorithm is benchmarked against principal component analysis in experiments on images of handwritten digits.