• Journal of Welding and Joining
• /
• v.13 no.2
• /
• pp.139-149
• /
• 1995

Many pressure vessels for the power plants are fabricated from low alloy ferritic steels. The inner sides of the pressure vessels are commonly weld_cladded with austenitic stainless steels to minimize problems of corrosive attack. The submerged-arc welding(SAW) process is now used in preference to other processes because of the possibilities open to automation to reduce the overaII welding times. The most reliable way to avoid underclad cracks(UCC) which are often detected at the overlap of the clad beads is to use nonsusceptible steels such as SA508 class 3. At present domestically developed forging steel of SA508 cl.S is now being cladded with single layer by using 90mm wide strip, which transfers higher heat input into the base metal compared to the conventional two layers strip cladding which has been in wide use with 30-60 mm wide strip. But the current indices for the influence of heat input on crack susceptibility are not accurate enough to express the subtle difference in crack susceptibility of the steel. Therefore, the purpose of this present study is: l) To determine UCC susceptibility on domestic forging steel, SA508 cl.S cladded with single layer by using submerged arc 90mm strip and, 2) To optimize heat input range by which the crack susceptibility could be eliminated.

• Journal of Information Technology Applications and Management
• /
• v.19 no.1
• /
• pp.1-12
• /
• 2012

Data mining is the process of searching and analyzing large quantities of data for finding out meaningful patterns and rules. Artificial Neural Network (ANN) is one of the tools of data mining which is becoming very popular in forecasting the future values. Some of the areas where it is used are banking, medicine, retailing and fraud detection. In finance, artificial neural network is used in various disciplines including stock market forecasting. In the stock market time series, due to high volatility, it is very important to choose a model which reads volatility and forecasts the future values considering volatility as one of the major attributes for forecasting. In this paper, an attempt is made to develop two models - one using feed forward back propagation Artificial Neural Network and the other using Autoregressive Conditional Heteroskedasticity (ARCH) technique for forecasting stock market returns. Various parameters which are considered for the design of optimal ANN model development are input and output data normalization, transfer function and neuron/s at input, hidden and output layers, number of hidden layers, values with respect to momentum, learning rate and error tolerance. Simulations have been done using prices of daily close of Sensex. Stock market returns are chosen as input data and output is the forecasted return. Simulations of the Model have been done using MATLAB$^{(R)}$ 6.1.0.450 and EViews 4.1. Convergence and performance of models have been evaluated on the basis of the simulation results. Performance evaluation is done on the basis of the errors calculated between the actual and predicted values.

• Journal of Korean Society of Steel Construction
• /
• v.10 no.4 s.37
• /
• pp.709-720
• /
• 1998

In order to study the distribution of welding residual stress through the plate thickness. experiment and analysis of fillet welding details were carried out. Especially, a residual stress in the weld root part of T-joint fillet weld whose measurement was difficult up to now was measured. By using the heat input and the number of the weld layers as parameters, the distribution of the 3-dimensional residual stress was investigated. As a result, we can say that with increasing the heat input, the residual stress in the weld toe and weld root barely changes. But, the area of the tensile residual stress became wide. Then, comparing a single pass with multi-pass weld method, it was found that the residual stress decreased more in multi-pass than in single pass. Moreover, it was found the thing that the area of tensile residual stress by multi-pass is lower than that by single-pass in the near weld part.

• Structural Engineering and Mechanics
• /
• v.77 no.1
• /
• pp.47-56
• /
• 2021

This paper deals with damage detection using a Gapped Smoothing Method (GSM) combined with deep learning. Convolutional Neural Network (CNN) is a model of deep learning. CNN has an input layer, an output layer, and a number of hidden layers that consist of convolutional layers. The input layer is a tensor with shape (number of images) × (image width) × (image height) × (image depth). An activation function is applied each time to this tensor passing through a hidden layer and the last layer is the fully connected layer. After the fully connected layer, the output layer, which is the final layer, is predicted by CNN. In this paper, a complete machine learning system is introduced. The training data was taken from a Finite Element (FE) model. The input images are the contour plots of curvature gapped smooth damage index. A free-free beam is used as a case study. In the first step, the FE model of the beam was used to generate data. The collected data were then divided into two parts, i.e. 70% for training and 30% for validation. In the second step, the proposed CNN was trained using training data and then validated using available data. Furthermore, a vibration experiment on steel damaged beam in free-free support condition was carried out in the laboratory to test the method. A total number of 15 accelerometers were set up to measure the mode shapes and calculate the curvature gapped smooth of the damaged beam. Two scenarios were introduced with different severities of the damage. The results showed that the trained CNN was successful in detecting the location as well as the severity of the damage in the experimental damaged beam.

• International journal of advanced smart convergence
• /
• v.12 no.2
• /
• pp.25-33
• /
• 2023

The purpose of this study is to evaluate the impact of intermediate features on FER performance. To achieve this objective, intermediate features were extracted from the input images at specific layers (FM1~FM4) of the pre-trained network (Resnet-18). These extracted intermediate features and original images were used as inputs to the vision transformer (ViT), and the FER performance was compared. As a result, when using a single image as input, using intermediate features extracted from FM2 yielded the best performance (training accuracy: 94.35%, testing accuracy: 75.51%). When using the original image as input, the training accuracy was 91.32% and the testing accuracy was 74.68%. However, when combining the original image with intermediate features as input, the best FER performance was achieved by combining the original image with FM2, FM3, and FM4 (training accuracy: 97.88%, testing accuracy: 79.21%). These results imply that incorporating intermediate features alongside the original image can lead to superior performance. The findings can be referenced and utilized when designing the preprocessing stages of a deep learning model in FER. By considering the effectiveness of using intermediate features, practitioners can make informed decisions to enhance the performance of FER systems.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.9 no.7
• /
• pp.1452-1457
• /
• 2005

In the QR Decomposition-based (QRD) decoding class, the system performance is sensitive to the error propagation. Thus, it is critical to correctly decode the previous layers. One apprach to desensitize the error propagation is to propose the optimal decoding order of layers. In this wort we propose a new extended-list Soled QRD-based (SQRD) decoding approach. In the proposed decoding scheme, the solution of the few first layers is extended as the list of promising possible solutions. By doing so, the diversity of the lowest layer is increased. As a result, the system performance is less sensitive to the error propagation than its counterparts. The proposed approach is verified by the computer simulation results.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 1998.11a
• /
• pp.73-76
• /
• 1998

We give pressure stimulation into organic thin films and detect the induced displacement current. then manufacture a device under the accumulation condition that the state surface pressure is 15[mN/m]. In processing of a device manufacture. We can see the process is good from the change of a surface pressure for organic thin films and transfer ratio of area per molecule. The structure of manufactured device is Au/organic thin films(polyimide)/Au, the number of accumulated layers are 31,35, and 41. I-V characteristic of the device is measured from 0[V] to +5[V]. The maximum value of measured current is increased as the number of accumulated layers are decreased. The resistance for the number of accumulated layers, the energy density for an input voltage show desired results, and the insulation of a thin film is better as the interval between electrodes is larger.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.8
• /
• pp.2098-2115
• /
• 1999

One of the unsolved problems in Neural Networks is the interpretation of hidden layers. This paper defines the Core-Net which has an input(p levels) and an output(q levels) with 2-layers as a basic circuit of neural network. In have suggested an equation, {{{{ {a}_{p,q} = {{q}^{2}} over {2}p(p-1)- { q} over {2 } (3 { p}^{2 } -7p+2)+ { p}^{2 }-3p+2}}}}, whichs ws the capacity of the Core-Net and have proved it by using the mathematical induction. It has been also shown that some of the problems with hidden layers can be solved by using the Core-Net and using simulation of an example.

• ETRI Journal
• /
• v.42 no.2
• /
• pp.230-238
• /
• 2020

Human-object interaction (HOI) detection is a popular computer vision task that detects interactions between humans and objects. This task can be useful in many applications that require a deeper understanding of semantic scenes. Current HOI detection networks typically consist of a feature extractor followed by detection layers comprising small filters (eg, 1 × 1 or 3 × 3). Although small filters can capture local spatial features with a few parameters, they fail to capture larger context information relevant for recognizing interactions between humans and distant objects owing to their small receptive regions. Hence, we herein propose a three-stream HOI detection network that employs a context convolution module (CCM) in each stream branch. The CCM can capture larger contexts from input feature maps by adopting combinations of large separable convolution layers and residual-based convolution layers without increasing the number of parameters by using fewer large separable filters. We evaluate our HOI detection method using two benchmark datasets, V-COCO and HICO-DET, and demonstrate its state-of-the-art performance.

• Korean Journal of Computational Design and Engineering
• /
• v.7 no.3
• /
• pp.148-156
• /
• 2002

Most Rapid Prototyping (RP) processes adopt a solid Computer Aided Design (CAD) model, which will be sliced into thin layers of constant thickness in the building direction. Each cross-sectional layer is successively deposited and, simultaneously, bonded onto the previous layer; and eventually the stacked layers from a physical part of the model. A new RP process, the transfer-type Variable Lamination Manufacturing process using expandable polystyrene foam sheet (VLM-ST), has been developed to reduce building time and to improve the surface finish of parts with the thick layers and a sloping surface. This paper describes the generation of Unit Shape Layer (USL), the cutting path data of the linen. hotwire cutter for the VLM-ST process. USL is a three-dimensional layer with a thickness of more than 1 mm and a side slope, and it is the basic unit of cutting and building in the VLM-ST process. USL includes data such as layer thickness, positional coordinates, side angles of each layer, hotwire cutting speed, the heat input to the hotwire, and reference shape. The procedure of generating USL is as follows: (1)Generation of the mid-slice from the CAD model, (2)Conversion of the mid-slice into a simply connected domain, (3)Generation to the reference shape for the mid-slice, (4)Calculation of the rotation angle of the hotwire of the cutting system.