• 전자공학회논문지CI
• /
• 제49권2호
• /
• pp.83-89
• /
• 2012

지구의 중요한 천연자원인 산림을 포함한 자연 식생환경은 지난 1세기 동안 많은 변화를 겪으며 기후에도 영향을 미치게 되어 현재 지구적 차원의 관심 속에서 다양한 연구가 진행되고 있다. 원격탐사는 분광적 특성을 이용하여 식생의 특성을 탐지할 수 있어 식생자원을 모니터링하는데 매우 효율적인 수단이다. 이러한 연구에서는 보통 원격탐사 측정을 분석하여 관찰된 화소가 식생을 포함하고 있는 정도를 나타내는 식생지수가 사용되고 있는데 NDVI가 이중 가장 많이 사용되는 식생지수이다. 본 논문에서는 MODIS NDVI 시계열 자료를 이용하여 자동으로 식생의 변화를 탐지해 가는 방법론이 제안되어 있다. 변화탐지를 위해 비모수 방법의 신경망 모형이 사용되었고 특성벡터로는 한 화소에서 다중 시기의 NDVI 차이와 더불어 NDVI 시계열 자료의 시간상의 관계가 함께 고려될수 있도록 제안되었다. 사용된 모형의 테스트를 위해 2006년부터 2011년까지 한반도 지역에 대한 MODIS MYD13Q1 자료가 사용되었다.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제29권1_2호
• /
• pp.53-64
• /
• 2002

연관규칙 탐사기법은 트랜잭션들을 대상으로 항목간 또는 속성간의 연관관계를 발견하는 방법으로, 데이터 집합의 구조를 쉽게 통찰할 수 있다는 장점으로 인하여 활발히 연구되어 왔다. 그러나 현재까지의 연구들은 전체 사용자 중 공통적인 특성을 지닌 사용자 그룹이 존재할 경우, 이러한 그룹별 연관규칙을 찾아낼 수 없다는 한계점을 지닌다. 본 논문에서는 이러한 점을 해결하기 위하여, 속성선택 및 사용자 구분 기법을 이용하여 사용자를 부분집합으로 구분하고 그 부분집합별로 연관규칙을 발견한다. 또한 위와 같이 얻어진 지역적 연관규칙이 전체 사용자를 대상으로 한 전역적 연관규칙보다 해당 부분집합에 더욱 적합하다는 사실을 여러 연관규칙 평가치를 이용하여 평가한다.

• 한국재료학회지
• /
• 제18권2호
• /
• pp.73-76
• /
• 2008

[ $SF_6$ ] gas has been widely used as an insulating, cleaning and covering gas due to its outstanding insulating feature and because of its inert properties. However, the global warming potential of $SF_6$ gas is extremely high relative to typical global warming gases such as $CO_2$, CFCs, and $CH_4$. For these reasons, it is necessary to separate and collect waste $SF_6$ gas. In this study, the effects of a surfactant (Tween) on the formation rate of $SF_6$ gas hydrates were investigated. The $SF_6$ gas hydrate formation rate increased with the addition of Tween and showed a nearly 6.5 times faster hydrate formation rate with an addition of 0.2 wt.% Tween compared to an addition of pure water. This is believed to be due to the increased solubility of $SF_6$ gas with the addition of the surfactant. It was also found that $SF_6$ gas hydrate in the surfactant solution showed two-stage hydrate formation rates with a formation rate that increased rapidly in the 2nd stage.

• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제6권1호
• /
• pp.103-112
• /
• 2000

지능망과 이동 통신 시장의 활성화로 차세대 지능망 서비스를 구현함에 있어 이동 서비스 가입자 및 이동 서비스 이용자 등의 서비스 수요가 더욱 증가될 것이 고려되어야 한다. 따라서, 본 논문에서는 가장 많이 사용되는 지능망 서비스중의 하나인 착신 과금 서비스에 대량호를 처리할 수 있는 호 대기 기능을 추가하여 호 성공률을 높이고, 서비스 가입자의 범위를 기존의 유선망 가입자뿐만 아니라 이동 단말을 소유한 무선망 가입자까지 확장하여 가입자의 이동성을 고려한 Global Service Logic을 설계한다. 또한 서비스에 가입된 각 가입자 단말기들의 위치 정보를 관리하여, 그들의 이동에 따른 동적 그룹핑을 수행하는 Queue Manager의 구조 및 동작 메커니즘을 설계하며, 호 시도율과 drop rate에 따라 필요한 큐의 적정 크기와 이에 대한 큐에서의 대기 시간을 제시한다.

• 한국컴퓨터정보학회논문지
• /
• 제11권3호
• /
• pp.31-39
• /
• 2006

본 논문에서는 지능로봇. 지능형자동차. 지능형빌딩 등에 다양하게 활용될 수 있는 3차원 환경과 여기에 포함된 물체의 실시간 인식을 위한 새로운 접근 방법을 제안한다. 본 논문에서는 먼저 사람이 환경을 인식하고 상호작용하는 데 사용하는 3가지 기본 원칙을 설정하고, 이 기본 원칙들을 이용하여 실시간 3차원 환경 및 물체 인식을 위한 통합된 방법을 제시한다. 이들 3가지 기본 원칙은 다음과 같다. 첫째, 전역 적인 평면 특징들을 인식함으로써 작업환경의 기하학적 구조에 대한 개략적 특성화를 고속으로 진행한다. 둘째, 작업환경 속에서 기존에 알려진 물체를 먼저 빠르게 인식하고 이를 데이터베이스 내에 저장되어 있는 물체의 모델로 교체한다. 셋째, 다중 해상도 Octree 표현 방법을 이용하여 기타 영역을 주어진 작업의 필요에 따라 적응적으로 실시간 모델링 한다. 본 논문에서는 3차원 SIFT로 언급되는 3차원 좌표를 가지는 SIFT특징들을 3차원 좌표정보와 함께 확장하여 사용함으로서 전역적 평면 특징의 빠른 추출, 고속의 물체 인식, 빠른 장면 정합 등의 기능에 활용하고 이와 동시에 스테레오 카메라로부터 입력되는 3차원 좌표의 잡음과 불완전성을 극복한다.

• 한국기후변화학회지
• /
• 제6권2호
• /
• pp.95-104
• /
• 2015

To discuss whether we have credible estimations about historical surface temperature evolution since industrial revolution or not, present study investigates consistencies and differences of averaged surface air temperature since 1900 between the multiple data sources: Hadley Center Climate Research Unit (HadCRU4) surface air temperature data, ECMWF 20 Century Reanalysis data (ERA20CR), and NCEP 20 Century Reanalysis data (NCEP20CR). Averaged surface temperatures are obtained for the global, polar (90S~60S, 60N~0N), midlatitude (60S~30S, 30N~60N), tropical (30S~30N) region, separately. From the analysis, we show that: 1) spatio-temporal inhomogenity and scarcity of HadCRU4 data are not major obstacles in the reliable estimation of global surface air temperature. 2) Globally averaged temperature variability is largely contributed by those of tropical and midlatitude, which occupy more than 70% of earth surface in area. 3) Both data show consistent temperature variability in tropical region. 4) ERA20CR does not capture warm period over Arctic region in early 1900s, which is obvious feature in HadCRU4 data. Discrepancies among datasets suggest that high-level caution is needed especially in the interpretation of large Arctic warming in the early 1900s, which is often regarded as a natural variability in the Arctic region.

• 한국지구과학회지
• /
• 제40권4호
• /
• pp.353-381
• /
• 2019

The general world model for homogeneous and isotropic universe has been proposed. For this purpose, we introduce a global and fiducial system of reference (world reference frame) constructed on a (4+1)-dimensional space-time, and assume that the universe is spatially a 3-dimensional hypersurface embedded in the 4-dimensional space. The simultaneity for the entire universe has been specified by the global time coordinate. We define the line element as the separation between two neighboring events on the expanding universe that are distinct in space and time, as viewed in the world reference frame. The information that determines the kinematics of the geometry of the universe such as size and expansion rate has been included in the new metric. The Einstein's field equations with the new metric imply that closed, flat, and open universes are filled with positive, zero, and negative energy, respectively. The curvature of the universe is determined by the sign of mean energy density. We have demonstrated that the flat universe is empty and stationary, equivalent to the Minkowski space-time, and that the universe with positive energy density is always spatially closed and finite. In the closed universe, the proper time of a comoving observer does not elapse uniformly as judged in the world reference frame, in which both cosmic expansion and time-varying light speeds cannot exceed the limiting speed of the special relativity. We have also reconstructed cosmic evolution histories of the closed world models that are consistent with recent astronomical observations, and derived useful formulas such as energy-momentum relation of particles, redshift, total energy in the universe, cosmic distance and time scales, and so forth. The notable feature of the spatially closed universe is that the universe started from a non-singular point in the sense that physical quantities have finite values at the initial time as judged in the world reference frame. It has also been shown that the inflation with positive acceleration at the earliest epoch is improbable.

• 산업경영시스템학회지
• /
• 제44권2호
• /
• pp.58-65
• /
• 2021

Meta-heuristic algorithms have been developed to efficiently solve difficult problems and obtain a global optimal solution. A common feature mimics phenomenon occurring in nature and reliably improves the solution through repetition. And at the same time, the probability is used to deviate from the regional optimal solution and approach the global optimal solution. This study compares the algorithm created based on the above common points with existed SA and HS to show advantages in time and accuracy of results. Existing algorithms have problems of low accuracy, high memory, long runtime, and ignorance. In a two-variable polynomial, the existing algorithms show that the memory increases and the accuracy decrease. In order to improve the accuracy, the new algorithm increases the number of initial inputs and increases the efficiency of the search by introducing a direction using vectors. And, in order to solve the optimization problem, the results of the last experiment were learned to show the learning effect in the next experiment. The new algorithm found a solution in a short time under the experimental conditions of long iteration counts using a two-variable polynomial and showed high accuracy. And, it shows that the learning effect is effective in repeated experiments.

• 로봇학회논문지
• /
• 제17권3호
• /
• pp.365-372
• /
• 2022

The accurate object recognition is important for the precise and accurate manipulation. To enhance the recognition performance, we can use various types of sensors. In general, acquired data from sensors have a high sampling rate. So, in the past, the RNN-based model is commonly used to handle and analyze the time-series sensor data. However, the RNN-based model has limitations of excessive parameters. CNN-based model also can be used to analyze time-series input data. However, CNN-based model also has limitations of the small receptive field in early layers. For this reason, when we use a CNN-based model, model architecture should be deeper and heavier to extract useful global features. Thus, traditional methods like RN N -based and CN N -based model needs huge amount of learning parameters. Recently studied result shows that Fast Fourier Convolution (FFC) can overcome the limitations of traditional methods. This operator can extract global features from the first hidden layer, so it can be effectively used for feature extracting of sensor data that have a high sampling rate. In this paper, we propose the algorithm to recognize objects using tactile sensor data and the FFC model. The data was acquired from 11 types of objects to verify our posed model. We collected pressure, current, position data when the gripper grasps the objects by random force. As a result, the accuracy is enhanced from 84.66% to 91.43% when we use the proposed FFC-based model instead of the traditional model.

• 한국멀티미디어학회논문지
• /
• 제25권5호
• /
• pp.748-756
• /
• 2022

Thanks to the competition of AlphaGo and Sedol Lee, machine learning has received world-wide attention and huge investments. The performance improvement of computing devices greatly contributed to big data processing and the development of neural networks. Artificial intelligence not only imitates human beings in many fields, but also seems to be better than human capabilities. Although humans' creation is still considered to be better and higher, several artificial intelligences continue to challenge human creativity. The quality of some creative outcomes by AI is as good as the real ones produced by human beings. Sometimes they are not distinguishable, because the neural network has the competence to learn the common features contained in big data and copy them. In order to confirm whether artificial intelligence can express the inherent characteristics of different arts, this paper proposes a new neural network model called Humming. It is an experimental model that combines vgg16, which extracts image features, and DeepJ's architecture, which excels in creating various genres of music. A dataset produced by our experiment shows meaningful and valid results. Different results, however, are produced when the amount of data is increased. The neural network produced a similar pattern of music even though it was a different classification of images, which was not what we were aiming for. However, these new attempts may have explicit significance as a starting point for feature transfer that will be further studied.