• Journal of Welding and Joining
• /
• v.13 no.3
• /
• pp.77-88
• /
• 1995

A prediction method for determining the welding residual stress by artificial neural network is proposed. A three-dimensional transient thermomechanical analysis has been performed for the CO$_{2}$ arc welding using the finite element method. The first part of numerical analysis performs a three-dimensional transient heat transfer analysis, and the second part then uses the results of the first part and performs a three-dimensional transient thermo-elastic-plastic analysis to compute transient and residual stresses in the weld. Data from the finite element method are used to train a backpropagation neural network to predict the residual stress. Architecturally, the fully interconnected network consists of an input layer for the voltage and current, a hidden layer to accommodate the ailure mechanism mapping, and an output layer for the residual stress. The trained network is then applied to the prediction of residual stress in the four specimens. It is concluded that the accuracy of the neural network predicting method is fully comparable with the accuracy achieved by the traditional predicting method.

• Proceedings of the KIEE Conference
• /
• 2006.04a
• /
• pp.120-121
• /
• 2006

In this paper, We have designed a face recognition system based on the embedded Linux. This paper has an aim in embedded system to recognize the face more exactly. At first, the contrast of the face image is adjusted with lightening compensation method, the skin and lip color is founded based on YCbCr values from the compensated image. To take advantage of the method based on feature and appearance, these methods are applied to the eyes which has the most highly recognition rate of all the part of the human face. For eyes detecting, which is the most important component of the face recognition, we calculate the horizontal gradient of the face image and the maximum value. This part of the face is resized for fitting the eye image. The image, which is resized for fit to the eye image stored to be compared, is extracted to be the feature vectors using the continuous wavelet transform and these vectors are decided to be whether the same person or not with PNN, to miminize the error rate, the accuracy is analyzed due to the rotation or movement of the face. Also last part of this paper we represent many cases to prove the algorithm contains the feature vector extraction and accuracy of the comparison method.

• Journal of Conservation Science
• /
• v.35 no.1
• /
• pp.71-80
• /
• 2019

This study examined the applicability of digital technologies based on three-dimensional(3D) scanning, modeling, and printing to the restoration of damaged artifacts. First, 3D close-range scanning was utilized to make a high-resolution polygon mesh model of a roof-end tile with a missing part, and a 3D virtual restoration of the missing part was conducted using a haptic interface. Furthermore, the virtual restoration model was printed out with a 3D printer using the material extrusion method and a PLA filament. Then, the additive structure of the printed output with a scanning electron microscope was observed and its shape accuracy was analyzed through 3D deviation analysis. It was discovered that the 3D printing output of the missing part has high dimensional accuracy and layer thickness, thus fitting extremely well with the fracture surface of the original roof-end tile. The convergence of digital virtual restoration based on 3D scanning and 3D printing technology has helped in minimizing contact with the artifact and broadening the choice of restoration materials significantly. In the future, if the efficiency of the virtual restoration modeling process is improved and the material stability of the printed output for the purpose of restoration is sufficiently verified, the usability of 3D digital technologies in cultural heritage restoration will increase.

• International Journal of Korean Welding Society
• /
• v.1 no.2
• /
• pp.51-60
• /
• 2001

A prediction method for determining the welding residual stress by artificial neural network is proposed. A three-dimensional transient thermo-mechanical analysis has been performed for the $CO_2$ arc welding using the finite element method. The first part of numerical analysis performs a three-dimensional transient heat transfer analysis, and the second part then uses the results of the first part and performs a three-dimensional transient thermo-elastic-plastic analysis to compute transient and residual stresses in the weld. Data from the finite element method are used to train a back propagation neural network to predict the residual stress. Architecturally, the fully interconnected network consists of an input layer for the voltage and current, a hidden layer to accommodate the failure mechanism mapping, and an output layer for the residual stress. The trained network is then applied to the prediction of residual stress in the four specimens. It is concluded that the accuracy of the neural network predicting method is fully comparable with the accuracy achieved by the traditional predicting method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.37.1-37
• /
• 2001

Electric plating is used in various industry field. Specially, pulse plating is able to deposit material at high current density compared to conventional DC plating. For example, pulse plating can get more fine grain, can improve adhesion and metal distribution and current efficiency, can reduce internal stress and crack. Therefore, we developed bidirection pulse power supply(BPPS) which has high speed pulse current and high current density and improve deposition quality and increase plating speed in this paper. BPPS(Bidirection pulse power supply) needs high speed rising time, falling time and output current accuracy. BPPS consists of rectifier part, chopper part, invertor part, and control part. Rectifier part changes outprt current direction.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.643-648
• /
• 2005

Generally, ships have marks of various shapes on outside of the hull. Among them, so called "Draft Mark" indicates the distance from the bottom of the keel to the waterline. Draft marks are used to determine the displacement and other properties of the ship for stability and control purposes. These marks are made up of welding bead or sticking the steel plate on outside of the hull. To improve the confidence level of the ship owner, quality and accuracy of the draft mark is very important. So the automatic mark welding robot is used to enable a high quality and accurate manufacturing line. To improve the system portability, the system is divided into two distinct parts, namely mechanical part and control part. Mechanical part is robust, a lightweight, and easy to dismantle. The control part consists of an in-house developed controller, which is based on embedded Linux. Also, the control part consists of power line communication module to ensure the applicability of the controller in manufacturing line. In this paper, the methodologies of control and configuration of the robot are discussed.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2002.05a
• /
• pp.399-404
• /
• 2002

Improvement of measuring accuracy is an essential part of quality control manufacturing process. The OMM is less than the CMM in measure accuracy but the OMM system is more efficient, easier to use than other measurement system. About 40~70% of the machine tool errors are induced by the thermal errors. Therefore, a key requirement for improving the measuring accuracy is to reduce the geometric and thermal errors. Thermal errors are measured by a ball bar system and predicted by the thermal error modeling. Furthermore, using the pre-defined thermal error map approach compensates the geometric accuracy of the OMM. Appropriate experiments are performed using ball-bar system, temperature measuring devices and touch-type probe. Compensated results are compared with those obtained using CMM to verify the proposed methods.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2002.05a
• /
• pp.546-550
• /
• 2002

This paper proposes an error compensation method that improves accuracy with geometry information of injection molding parts. Geometric information can give an improved accuracy in reverse engineering. Measuring data can not lead to get accurate geometric model, including errors of physical parts and measuring machines. Measuring data include errors which can be classified into two types. One is molding error in product, the other is measuring error. Measuring error includes optical error of laser scanner, deformation by probe forces of CMM and machine error. It is important to compensate these in reverse engineering. Least square method(LSM) provides the cloud data with a geometry compensation, improving accuracy of geometry. Also, the functional shape of a part and design concept can be reconstructed by error compensation using geometry information.

• Transactions of Materials Processing
• /
• v.23 no.6
• /
• pp.380-385
• /
• 2014

An analysis procedure is proposed to improve the prediction accuracy of springback as well as to evaluate the structural stability of the tooling used for fabricating a side sill part from UHSS. The analysis couples the stamping analysis and the subsequent analysis of the tool structural. The deformation and stress results for the tool structure are obtained from the proposed analysis procedure. The results show that the amount of deformation and stresses are so high that the tool structure must be reinforced and the tooling design must consider structural stability. Springback is predicted with CAE in order to compare the prediction accuracy between the given tool geometry and the geometry from the structural analysis. The simulation results with the deformed tool can predict the experimental springback tendency accurately.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.39 no.8
• /
• pp.856-867
• /
• 1990

In this paper, the convergence and the learning accuracy of the back-propagation (BP) method in neural network are investigated by 1) analyzing the reason for decelerating the convergence of BP method and examining the rapid deceleration of the convergence when the learning is executed on the part of sigmoid activation function with the very small first derivative and 2) proposing the modified logistic activation function by defining, the convergence factor based on the analysis. Learning on the output patterns of binary as well as analog forms are tested by the proposed method. In binary output patter, the test results show that the convergence is accelerated and the learning accuracy is improved, and the weights and thresholds are converged so that the stability of neural network can be enhanced. In analog output patter, the results show that with extensive initial transient phenomena the learning error is decreased according to the convergence factor, subsequently the learning accuracy is enhanced.