• 한국지능시스템학회논문지
• /
• 제15권6호
• /
• pp.682-687
• /
• 2005

본 논문에서는 생물의 2차원영상에서 5가지 특징을 추출한 다음 약품에 대한 생물의 행동 패턴 반응에 대하여 의사결정나무를 적용하여 패턴의 인식 및 분류를 하였다. 생물의 행동패턴을 대변하는 물리적인 특징인, 속도, 방향전환 각도, 이동거리에 대하여 각각 중간이상속도비율 FFT(Fast Fourier Transform), 2차원 정사영 면적, 프렉탈 차원, 무게중심을 사용하여 특징을 추출하였다. 이렇게 추출된 5가지의 특징변수들을 사용하여 의사결정나무 모델을 구성한 다음 생물의 약품 첨가에 대한 반응을 분석하였다 또한 결과에서는 기존의 생물의 행동패턴 구분에 쓰였던 전형적인 기법(conventional methods) 보다 본 연구에서 적용한 의사결정나무가 생물의 행동패턴이 가지는 물리적 요소에 대한 독해력을 가짐을 보임으로써 특정 환경에서 이동행동에 대한 분석을 용이하게 하고자 하였다.

• Journal of Computing Science and Engineering
• /
• 제11권3호
• /
• pp.79-91
• /
• 2017

In this paper, the abdominal-deformation measurement scheme is conducted on a shape-flexible mannequin using the DIC technique in a stereo-vision system. Firstly, during the integer-pixel displacement search, a novel fractal dimension based on an adaptive-ellipse subset area is developed to track an integer pixel between the reference and deformed images. Secondly, at the subpixel registration, a new mutual-learning adaptive particle swarm optimization (MLADPSO) algorithm is employed to locate the subpixel precisely. Dynamic adjustments of the particle flight velocities that are according to the deformation extent of each interest point are utilized for enhancing the accuracy of the subpixel registration. A test is performed on the abdominal-deformation measurement of the shape-flexible mannequin. The experiment results indicate that under the guarantee of its measurement accuracy without the cause of any loss, the time-consumption of the proposed scheme is significantly more efficient than that of the conventional method, particularly in the case of a large number of interest points.

• 상하수도학회지
• /
• 제23권4호
• /
• pp.397-406
• /
• 2009

This study was conducted to monitor the floc sizes forming in the mixing zone in the water treatment plant. The dosing amount of poly aluminium chloride(PAC) was determined by particle dispersion analyzer(iPDA) in the lab and field scale test. During a field test period, PAC coagulant was used and the raw water was taken from Nakdong river. PAC wad diluted to activate the coagulant, leading to bring the more homogeneous dispersion in the shorter time. To monitor the floc sizes, the unit of floc size index(FSI) was used. With increasing of raw water turbidity, FSI value was increased. Also, the increased dosing amount of PAC brought the increased FSI and with overdosing of coagulant was in turn decreased. When the PAC was fed into the raw water after dilution in a field scale test, the width of FSI was narrower compared with the feeding of the mother liquor of PAC, implying that the formed flocs are denser and more uniform sizes. The width of FSI in average was varied on depending on the basicity of coagulant. Also, dF value, fractal dimension was evalued with the different coagulants, showing from 2.01 to 2.03. On-line floc monitor was effective for the optimal dosing in the drinking water plant.

• Journal of Mechanical Science and Technology
• /
• 제19권7호
• /
• pp.1393-1404
• /
• 2005

Fatigue crack growth and life have been estimated based on established empirical equations. In this paper, an alternative method using artificial neural network (ANN) -based model developed to predict fatigue damages simultaneously. To learn and generalize the ANN, fatigue crack growth rate and life data were built up using in-plane bending fatigue test results. Single fracture mechanical parameter or nondestructive parameter can't predict fatigue damage accurately but multiple fracture mechanical parameters or nondestructive parameters can. Existing fatigue damage modeling used this merit but limited real-time damage monitoring. Therefore, this study shows fatigue damage model using backpropagation neural networks on the basis of X -ray half breadth ratio B / $B_o$, fractal dimension $D_f$ and fracture mechanical parameters can estimate fatigue crack growth rate da/ dN and cycle ratio N / $N_f$ at the same time within engineering limit error ($5\%$).

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 1999년도 추계학술대회-지능형 정보기술과 미래조직 Information Technology and Future Organization
• /
• pp.365-373
• /
• 1999

Recently, numerous studies have demonstrated that artificial intelligence such as neural networks can be an alternative methodology for classification problems to which traditional statistical methods have long been applied. In building neural network model, the selection of independent and dependent variables should be approached with great care and should be treated as a model construction process. Irrespective of the efficiency of a learning procedure in terms of convergence, generalization and stability, the ultimate performance of the estimator will depend on the relevance of the selected input variables and the quality of the data used. Approaches developed in statistical methods such as correlation analysis and stepwise selection method are often very useful. These methods, however, may not be the optimal ones for the development of neural network models. In this paper, we propose a genetic algorithms approach to find an optimal or near optimal input variables for neural network modeling. The proposed approach is demonstrated by applications to bankruptcy prediction modeling. Our experimental results show that this approach increases overall classification accuracy rate significantly.

• 대한기계학회논문집A
• /
• 제24권3호
• /
• pp.634-644
• /
• 2000

Fatigue crack growth and life is estimated by various fracture mechanical parameters but affected by load, material and environment. Fatigue character of component without surface notch cannot be e valuated by above-mentioned parameters due to microstructure of in-service material. Single fracture mechanical parameter or nondestructive parameter cannot predict fatigue damage in arbitrary boundary condition but multiple fracture mechanical parameters or nondestructive parameters can Fatigue crack growth modelling with three point representation scheme uses this merit but has limit on real-time monitoring. Therefore, this study shows fatigue damage model using backpropagatior. neural networks on the basis of X-ray half breadth ratio B/$B_o$ fractal dimension $D_f$ and fracture mechanical parameters can predict fatigue crack growth rate da/dN and cycle ratioN/$N_f$ at the same time within engineering estimated mean error(5%).

• Tribology and Lubricants
• /
• 제34권1호
• /
• pp.1-8
• /
• 2018

This paper presents the analysis of friction master curves for a sliding elastomer on rough granite. The hysteresis friction is calculated using an analytical model that considers the energy spent during the local deformation of the rubber due to surface asperities. The adhesion friction is also considered for dry friction prediction. The viscoelastic modulus of the rubber compound and the large-strain effective modulus are obtained from dynamic mechanical analysis (DMA). We accurately demonstrate the large strain of rubber that contacts with road substrate using the GW theory. We found that the rubber block deforms approximately to 40% strain. In addition, the viscoelastic master curve considering nonlinearity (at 40% strain) is derived based on the above finding. As viscoelasticity strongly depends on temperature, it can be assumed that the influence of velocity on friction is connected to the viscoelastic shift factors gained from DMA using the time-temperature superposition. In this study, we apply these shift factors to measure friction on dry granite over a velocity range for various temperatures. The measurements are compared to simulated hysteresis and adhesion friction using the Kluppel friction theory. Although friction results in the low-speed band match well with the simulation results, there are differences in the predicted and experimental results as the velocity increases. Thus, additional research is required for a more precise explanation of the viscoelastic material properties for better prediction of rubber friction characteristics.

• 한국지능시스템학회논문지
• /
• 제11권4호
• /
• pp.315-319
• /
• 2001

회전체 베어링 상태진단에 신뢰성을 갖기 위하여 여러 가지 진단 방법이 연구되고 있으며, 이때 이용하는 변수는 온도와 소음, 진동 그리고 윤활유가 있으며 분석 방법으로는 온도추이분석, 소음분석, 진동분석, 윤활제 분석방법이 주로 이용되고 있다. 본 연구에서는 압연기 베어링의 상태진단 변수로 베어링의 진동 신호를 선택하고 이 진동신호에서 비선형성이 강한 신호 즉 카오스적 거동이 있음을 정성적인 방법으로 타켄스의 매립법에 의한 상태공간 재구성과 포엔카레 단면, FFT, 히스토그램을 이용하고, 정량적인 방법으로 프랙탈 차원, 리아프노프 지수를 이용하여 확인하였다.

• 터널과지하공간
• /
• 제16권1호
• /
• pp.26-37
• /
• 2006

본 연구는 절리암반의 통계적 동질성에 대한 정량적 기준으로 절리의 평면밀도 및 길이를 동시에 고려할 수 있는 통계적 모수인 박스프랙털$(D_B)$의 적용성에 대해 논하였다. 길이분포와 평면밀도를 달리하는 총 129개의 절리연결망에서 박스 집계법을 이용하여 절리의 길이분포와 평면밀도의 변화에 따른 $D_B$의 변화특성을 고찰한 결과 $D_B$는 절리의 방향 또는 절리길이의 표준편차 변화에 영향을 받지 않고 전체절리에 대한 평균 절리길이와 평면밀도에만 영향을 받는다는 사실을 검증하였다. 또한 $D_B$는 절리의 길이와 평면밀도의 함수로써 공학적 지구조구 구분의 정량적 척도로 활용될 수 있음을 입증하였다. 본 연구의 현장 적용성을 검토하기 위하여 도로사면 및 지하구조물에서 박스 집계법을 적용한 사례연구를 수행하였다. 통계적 동질구역 구분에 있어서 일반적인 지질조건와 더불어 기존의 분할표 해석과 본 연구의 방법론을 병행하면 절리의 방향성, 평균길이 및 평면밀도가 종합적으로 고려된 공학적 지구조구의 구분이 가능할 것으로 사료된다.

• Geomechanics and Engineering
• /
• 제23권5호
• /
• pp.467-480
• /
• 2020

The failure behavior and tensile strength of sandstone materials under different strain rates are greatly different, especially under static loads and impact loads. In order to clearly investigate the failure mechanism of sandstone materials under static and impact loads, a series of Brazilian disc samples were used by employing green sandstone, red sandstone and black sandstone to carry out static and impact loading splitting tensile tests, and the failure properties subjected to two different loading conditions were analyzed and discussed. Subsequently, the failure behavior of sandstone materials also were simulated by finite element code. The good agreement between simulation results and experimental results can obtain the following significantly conclusions: (1) The relationship of the tensile strength among sandstone materials is that green sandstone < red sandstone < black sandstone, and the variation of the tensile sensitivity of sandstone materials is that green sandstone > red sandstone > black sandstone; (2) The mainly cause for the difference of dynamic tensile strength of sandstone materials is that the strength of crystal particles in sandstone material, and the tensile strength of sandstone is proportional to the fractal dimension; (3) The dynamic failure behavior of sandstone is greatly different from that of static failure behavior, and the dynamic tensile failure rate in dynamic failure behavior is about 54.92%.