• Computers and Concrete
• /
• 제20권1호
• /
• pp.111-118
• /
• 2017

The present study established prediction models based on multiple nonlinear regressions (MNLRs) and backpropagation neural networks (BPNs) for the expansion of cement mortar caused by oxidization slag that was used as a replacement of the aggregate. The data used for the models were obtained from actual laboratory tests on specimens that were produced with water/cement ratios of 0.485 or 1.5, within which 0%, 10%, 20%, 30%, 40%, or 50% of the cement had been replaced by oxidization slag from electric-arc furnaces; the samples underwent high-temperature curing at either $80^{\circ}C$ or $100^{\circ}C$ for 1-4 days. The varied mixing ratios, curing conditions, and water/cement ratios were all used as input parameters for the expansion prediction models, which were subsequently evaluated based on their performance levels. Models of both the MNLR and BPN groups exhibited $R^2$ values greater than 0.8, indicating the effectiveness of both models. However, the BPN models were found to be the most accurate models.

• International Journal of Concrete Structures and Materials
• /
• 제6권3호
• /
• pp.177-186
• /
• 2012

The temperature distributions of concrete structures strongly depend on the value of thermal conductivity of concrete. However, the thermal conductivity of concrete varies according to the composition of the constituents and the temperature and moisture conditions of concrete, which cause difficulty in accurately predicting the thermal conductivity value in concrete. For this reason, in this study, back-propagation neural network models on the basis of experimental values carried out by previous researchers have been utilized to effectively account for the influence of these variables. The neural networks were trained by 124 data sets with eleven parameters: nine concrete composition parameters (the ratio of water-cement, the percentage of fine and coarse aggregate, and the unit weight of water, cement, fine aggregate, coarse aggregate, fly ash and silica fume) and two concrete state parameters (the temperature and water content of concrete). Finally, the trained neural network models were evaluated by applying to other 28 measured values not included in the training of the neural networks. The result indicated that the proposed method using a back-propagation neural algorithm was effective at predicting the thermal conductivity of concrete.

• 대한기계학회논문집A
• /
• 제21권5호
• /
• pp.828-834
• /
• 1997

The knowledge of material stress-strain behavior is an essential requirement for design and analysis of deformation processes. Empirical stress-strain relationship and constitutive equations describing material behavior during deformation are being widely used, despite suffering some drawbacks in terms of ease of development, accuracy and speed. In the present study, back-propagation neural networks are used to model and predict the flow stresses of a HSLA steel under conditions of constant strain, strain rate and temperature. The performance of the network model is comparedto those of statistical models on rate equations. Well-trained network model provides fast and accurate results, making it superior to statistical models.

• 한국소음진동공학회논문집
• /
• 제19권4호
• /
• pp.425-432
• /
• 2009

Recently, reduction of road traffic noise in residential buildings has become one of the most important subjects. To reduce the road traffic noise, noise impact assessment by the road traffic prediction model is required before building construction. For reasonable road traffic noise prediction, it is required to analysis of various factors in road traffic prediction models. This paper was studied the road traffic noise propagation factors such as distance from road to building, receiver height, alignment angle of building and reflection coefficient of the building facade by two calculation models, RLS-90 and CRTN. The result showed that noise reduction was generally higher at bottom stories by ground absorption effect. The reflection coefficient of the building facade was affect of additional sound pressure level by facade reflecting. And alignment angle of building at $90^{\circ}$ was performed effective noise reduction better than $0^{\circ}$.

• Nuclear Engineering and Technology
• /
• 제52권2호
• /
• pp.287-295
• /
• 2020

Group method of data handling (GMDH) is considered one of the earliest deep learning methods. Deep learning gained additional interest in today's applications due to its capability to handle complex and high dimensional problems. In this study, multi-layer GMDH networks are used to perform uncertainty quantification (UQ) and sensitivity analysis (SA) of nuclear reactor simulations. GMDH is utilized as a surrogate/metamodel to replace high fidelity computer models with cheap-to-evaluate surrogate models, which facilitate UQ and SA tasks (e.g. variance decomposition, uncertainty propagation, etc.). GMDH performance is validated through two UQ applications in reactor simulations: (1) low dimensional input space (two-phase flow in a reactor channel), and (2) high dimensional space (8-group homogenized cross-sections). In both applications, GMDH networks show very good performance with small mean absolute and squared errors as well as high accuracy in capturing the target variance. GMDH is utilized afterward to perform UQ tasks such as variance decomposition through Sobol indices, and GMDH-based uncertainty propagation with large number of samples. GMDH performance is also compared to other surrogates including Gaussian processes and polynomial chaos expansions. The comparison shows that GMDH has competitive performance with the other methods for the low dimensional problem, and reliable performance for the high dimensional problem.

• International Journal of Control, Automation, and Systems
• /
• 제2권2호
• /
• pp.256-262
• /
• 2004

Conventionally, profile non-uniformity has been characterized by relying on approximated profile with angle or anisotropy. In this study, a new non-uniformity model for etch profile is presented by applying a discrete wavelet to the image obtained from a scanning electron microscopy (SEM). Prediction models for wavelet-transformed data are then constructed using a back-propagation neural network. The proposed method was applied to the data collected from the etching of tungsten material. Additionally, 7 experiments were conducted to obtain test data. Model performance was evaluated in terms of the average prediction accuracy (APA) and the best prediction accuracy (BPA). To take into account randomness in initial weights, two hundred models were generated for a given set of training factors. Behaviors of the APA and BPA were investigated as a function of training factors, including training tolerance, hidden neuron, initial weight distribution, and two slopes for bipolar sig-moid and linear function. For all variations in training factors, the APA was not consistent with the BPA. The prediction accuracy was optimized using three approaches, the best model based approach, the average model based approach and the combined model based approach. Despite the largest APA of the first approach, its BPA was smallest compared to the other two approaches.

• 자원환경지질
• /
• 제27권5호
• /
• pp.499-505
• /
• 1994

기존 미소결함에서의 국소적 응력집중은 새로운 미시적 손상의 원인이 되고, 이러한 미시적 손상은 또한 거시적 손상으로 발달하게 된다. 이들 미시적 손상에서 거시적 손상으로의 바달은 그 암석 및 암반의 변형특성으로 나타난다. 지금까지 응력하에서의 미소크랙의 거동에 대한 연구는 많이 되어왔으나, 실제암석의 파괴전 상태에서 미소크랙거동에 대한 역학적 해석은 아직 미비한 실정이다. 본 연구에서는 새로이 개발한 시험장치에 의한 정밀한 관찰로 손상 발달에 대한 이해를 더하였으며, 수학적 균질화 이론에 의해 수치해석 함으로서 그 역학성을 검토하였다.

• 한국우주과학회:학술대회논문집(한국우주과학회보)
• /
• 한국우주과학회 2009년도 한국우주과학회보 제18권2호
• /
• pp.33.2-33.2
• /
• 2009

Diffuse ionized gas (DIG or warm ionized medium, WIM) outside traditional regions is a major component of the interstellar medium (ISM) not only in our Galaxy, but also in other galaxies. It is generally believed that major fraction of the Halpha emission in the DIG is provided by OB stars. In the "standard" photoionization models, the Lyman continuum photons escaping from bright H II regions is the dominant source responsible for ionizing the DIG. Then, a complex density structure must provide the low-density paths that allow the photons to traverse kiloparsec scales and ionize the gas far from the OB stars not only at large heights above the midplane, but also within a galactic plane. Here, I present Monte-Carlo models to examine the propagation of the ionizing radiation leaked out of traditional H II regions into the diffuse ISM applied to two face-on spirals M 51 and NGC 7424. We find that the "standard" scenario requires absorption too unrealistically small to be believed, but the obtained scale-height of the galactic disk is consistent with those of edge-on galaxies. We also report that the probability density functions of the Halpha intensities of the DIG and H II regions in the galaxies are log-normal, indicating the turbulence property of the ISM.

• 전자공학회논문지T
• /
• 제35T권1호
• /
• pp.125-132
• /
• 1998

본 논문은 PCS 등 최근 급격히 활용도가 증가하고 있는 L-Band의 전파 전달 특성을 보편적인 국내 환경에서 실험하여 기지국 치국시 실질적으로 활용할 수 있는 방법을 모색하였다. 실험 전파는 1.805 ㎓에서 수직 편파를 사용하였고 예측 모델은 3㎞, 300m, 30m 대의 3가지 측정 구간을 설정하여 송수신 고정 상태에서 각각 3∼4개월씩 측정하였다. 예측 모델을 통한 실험에서 측정값은 환경에 따라 이론값과 1㏈∼6㏈ 정도의 차이가 있음을 알았고, 이 같은 내용이 기지국 설치 및 설계에 활용 가능한 자료로 적용될 수 있을 것으로 사료된다.

• Structural Engineering and Mechanics
• /
• 제36권4호
• /
• pp.481-497
• /
• 2010

The shaking table tests were conducted on two small-scale models (Model 1 and Model 2) to examine the earthquake-induced damage of a concrete gravity dam, which has been planned for the construction with the recommendation of the peak ground acceleration of the maximum credible earthquake of 0.42 g. This study deals with the numerical simulation of shaking table tests for two smallscale dam models. The plastic damage constitutive model is used to simulate the crack/damage behavior of the bentonite-concrete mixture material. The numerical results of the maximum failure acceleration and the crack/damage propagation are compared with experimental results. Numerical results of Model 1 showed similar crack/damage propagation pattern with experimental results, while for Model 2 the similar pattern was obtained by considering the modulus of elasticity of the first and second natural frequencies. The crack/damage initiated at the changing point in the downstream side and then propagated toward the upstream side. Crack/damage accumulation occurred in the neck area at acceleration amplitudes of around 0.55 g~0.60 g and 0.65 g~0.675 g for Model 1 and Model 2, respectively.