• 한국지반공학회:학술대회논문집
• /
• 한국지반공학회 2006년도 춘계 학술발표회 논문집
• /
• pp.1159-1164
• /
• 2006

In this paper an artificial neural network model is developed to estimate the unconfined compression strength of Deep Cement Mixing(DCM) treated soil. A database which consists of a number of unconfined compression test result compiled from 9 clay sites is used to train and test of the artificial neural network model. Developed neural network model requires water content of soil, unit weight of soil, passing percent of #200 sieve, weight of cement, w-c ratio as input variables. It is found that the developed artificial neural network model can predict more precise and reliable unconfined compression strength than the conventional empirical models.

• International Journal of Computer Science & Network Security
• /
• 제24권2호
• /
• pp.1-14
• /
• 2024

While text-to-image models have made remarkable progress in image synthesis, certain models, particularly generative diffusion models, have exhibited a noticeable bias to- wards generating images related to the culture of some developing countries. This paper introduces an empirical investigation aimed at mitigating the bias of image generative model. We achieve this by incorporating symbols representing Saudi culture into a stable diffusion model using the Dreambooth technique. CLIP score metric is used to assess the outcomes in this study. This paper also explores the impact of varying parameters for instance the quantity of training images and the learning rate. The findings reveal a substantial reduction in bias-related concerns and propose an innovative metric for evaluating cultural relevance.

• 한국습지학회지
• /
• 제4권1호
• /
• pp.51-61
• /
• 2002

The wetland is a biologically integrated system consisting of water, soil, bacteria, plants, and animals. The wetland helps sustain the ecosystem, control the micro-climate and flood, maintain the ground water level, and provide fishing grounds. From the environmental standpoint, the wetland plays a vital role in reducing water pollution by filtering out sand and other polluted matters, producing oxygen, absorbing chemicals and nutrients. For these reasons, interest in restoring the wetlands has been steadily increasing. Artificial wetland, which is also referred to as created wetland or constructed wetland, is an alternative to natural wetland. Like natural wetland, artificial wetland is environmentally friendly and can effectively lower pollutant levels. The Korea government is actively reviewing the construction of artificial wetlands in mining and water supply areas to decrease nonpoint pollutant sources. This paper attempts to develop a pollutant removal model for the water quality improvement function of artificial wetlands. Artificial wetland can improve the quality of the water; however, depending on the type of water inflow, vegetation and hydrology, its effect can be different.

• 지질공학
• /
• 제28권4호
• /
• pp.741-753
• /
• 2018

최근 우리나라에서 발생되는 국지적 가뭄은 지하수의 효율적 활용에 대한 관심을 증대시키고 있으며, 잉여의 물을 지층 내에 저장하는 지하수 인공함양 기술 도입의 필요성이 대두되고 있다. 본 연구에서는 충청남도내 퇴적 분지의 지하수 인공함양 대상지로의 가능성을 평가하기 위하여 1차 인자 3개, 2차 인자 7개로 구성된 AHP 모델을 개발하였으며, 10개 후보지에 적용한 결과를 토대로 인공신경망 모델을 구축하였다. AHP 모델은 후보지가 추가될 경우 수학적인 연산 과정에 의하여 최종 평가점수가 변하게 되나, 인공신경망 모델은 후보지별 고정적인 최종평가 점수를 제시하게 되어 인공함양 적지 선정 기준으로 사용할 수 있다. 충청남도 지역의 연구 결과, 인공신경망 모델의 최종 평가점수가 약 1.5점 이하인 경우에는 인공함양 후보지로서의 가능성이 낮은 것으로 평가되었다. 향후 타 지역에 대한 추가 연구 및 현장 조사를 통해 다양한 자료 군을 확보한다면 보다 보편적으로 적용할 수 있는 인공신경망 모델 도출이 가능할 것이다.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2005년도 추계학술대회 논문집
• /
• pp.244-247
• /
• 2005

Although several artificial disc designs have been developed for the treatment of discogenic low back pain and used clinically, biomechanical change with its implantation seldom studied. To evaluate the effect of artificial disc implantation on the biomechanics of lumbar spinal unit, nonlinear three-dimensional finite element model of L1-L5, S1 was developed and strain and stress of vertebral body and surrounding spinal ligaments were predicted. Intact osteoligamentous L1-L5, S1 model was created with 1-mm CT scan of a volunteer and known material property of each element were applied. This model also includes the effect of local muscles which was modeled with pre-strained spring elements. The intact model was validated with reported biomechanical data. Two models implanted with artificial discs, SB Charite or Prodisc, at L4/5 via anterior approach were also developed. The implanted model predictions were compared with that of intact model. Angular motion of vertebral body, force on spinal ligaments, facet joint contact force with $2\sim12$ Nm flexion-extension moment.

• 한국정밀공학회지
• /
• 제23권4호
• /
• pp.176-182
• /
• 2006

Although several artificial disc designs have been developed for the treatment of discogenic low back pain, biomechanical changes with its implantation were rarely studied. To evaluate the effect of artificial disc implantation on the biomechanics of functional spinal unit, a nonlinear three-dimensional finite element model of L4-L5 was developed with 1-mm CT scan data. Biomechanical analysis was performed for two different types of artificial disc having constrained and unconstrained instant center of rotation(ICR), ProDisc and SB Charite III model. The implanted model predictions were compared with that of intact model. Angular motion of vertebral body, forces on the spinal ligaments and facet joint, and stress distribution of vertebral endplate for flexion-extension, lateral bending, and axial rotation with a compressive preload of 400N were compared. The implanted model showed increased flexion-extension range of motion compared to that of intact model. Under 6Nm moment, the range of motion were 140%, 170% and 200% of intact in SB Charite III model and 133%, 137%, and 138% in ProDisc model. The increased stress distribution on vertebral endplate for implanted cases could be able to explain the heterotopic ossification around vertebral body in clinical observation. As a result of this study, it is obvious that implanted segment with artificial disc suffers from increased motion and stress that can result in accelerated degenerated change of surrounding structure. Unconstrained ICR model showed increased in motion but less stress in the implanted segment than constrained model.

• 한국지하수토양환경학회지:지하수토양환경
• /
• 제16권6호
• /
• pp.34-45
• /
• 2011

For the Hancheon drainage area in Jeju island, a groundwater flow model using Visual MODFLOW was developed to simulate artificial recharge through injection wells installed in the Hancheon reservoir. The model was used to analyze changes of the groundwater level and the water budget due to the artificial recharge. The model assumed that $2{\times}10^6m^3$ of storm water would recharge annually through the injection wells during the rainy season. The transient simulation results showed that the water level rose by 39.6 m at the nearest monitoring well and by 0.26 m at the well located 7 km downstream from the injection wells demonstrating a large extent of the affected area by the artificial recharge. It also shown that, at the time when the recharge ended in the 5th year, the water level increased by 81 m at the artificial reservoir and the radius of influence was about 2.1 km downstream toward the coast. The residence time of recharged groundwater was estimated to be no less than 5 years. The model also illustrated that 15 years of artificial recharge could increase the average linear velocity of groundwater up to 1540 m/yr, which showed 100 m/yr higher than before. Increase of groundwater storage due to artificial recharge was calculated to be $2.4{\times}10^6$ and $4.3{\times}10^6m^3$ at the end of the 5th and 10th years of artificial recharge, respectively. The rate of storage increase was gradually diminished afterwards, and storage increase of $5.0{\times}10^6m^3$ was retained after 15 years of artificial recharge. Conclusively, the artificial recharge system could augment $5.0{\times}10^6m^3$ of additional groundwater resources in the Hancheon area.

• 한국정보교육학회:학술대회논문집
• /
• 한국정보교육학회 2021년도 학술논문집
• /
• pp.349-354
• /
• 2021

본 연구에서는 2022년 개정 교육과정에서 소프트웨어와 인공지능 교육이 필수화되어 각 교과에서 인공지능과 융합할 수 있는 교육 모형을 제안한다. 제안한 인공지능 융합 교육 모형은 교과 내용(성취기준+주제)을 한 축으로 한다. 두 번째 축은 인공지능 도구이고, 세 번째 축은 인공지능 기술이고, 네 번째 축은 생활 속 적용 데이터이다. 인공지능을 각 교과에 적용하기 위해서 각 교과의 성취기준과 교과 내용에 인공지능 도구, 인공지능 기술, 생활 속 데이터 적용을 하여야 한다. 이렇게 성취기준과 교과 내용을 구성하면 각 교과와의 융합이 잘 된다고 볼 수 있다. 따라서 성취기준과 주제별로 교과서를 구성할 때에 인공지능 도구, 인공지능 기술, 생활속 데이터를 추가하는 것이 필요하다.

• 한국환경과학회지
• /
• 제9권6호
• /
• pp.455-462
• /
• 2000

This study aims at the development of the model for a forecasting of water quality in river basins using artificial neural network technique. Water quality by Artificial Neural Network Model forecasted and compared with observed values at the Sangju q and Dalsung stations in Nakdong river basin. For it, a multi-layer neural network was constructed to forecast river water quality. The neural network learns continuous-valued input and output data. Input data was selected as BOD, CO discharge and precipitation. As a result, it showed that method III of three methods was suitable more han other methods by statistical test(ME, MSE, Bias and VER). Therefore, it showed that Artificial Neural Network Model was suitable for forecasting river water quality.

• 정보교육학회논문지
• /
• 제21권1호
• /
• pp.139-149
• /
• 2017

21세기 지식 정보 사회에 인공지능 교육이 매우 중요하다. 4차 산업혁명 시대에 컴퓨터 교육에서 인공지능을 이해하고 컴퓨터 프로그래밍 교육을 해 보는 것이 매우 중요하지만 인공지능에 대해서 이해하고, 컴퓨터 프로그래밍 교육을 하는 교수 학습 모델이 없다. 본 연구에서는 제안하는 모델은 문제 이해 단계, 데이터 정리하기 단계, 인공지능 모델 정하기 단계, 프로그래밍하기 단계, 보고서 작성하기 단계로 구성된다. 프로그래밍하기 단계에서는 학생들의 수준에 적합하게 복사하기, 변형하기, 창조하기, 도전하기로 나눌 수 있다. 본 연구에서는 초등학교 교사들의 델파이 평가로 모델의 타당도를 입증하였고, 그에 따라 초등학생들이 쉽게 이해할 수 있는 사례를 만들었다. 본 연구의 결과는 초등학생들에서 인공지능 프로그램을 실습해 볼 수 있는 좋은 기회를 제공한다.