• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.3
• /
• pp.171-179
• /
• 2020

Recently, studies have shown that convolution neural networks are achieving the best performance in image classification, object detection, and image generation. Vision based defect inspection which is more economical than other defect inspection, is a very important for a factory automation. Although supervised anomaly detection algorithm has far exceeded the performance of traditional machine learning based method, it is inefficient for real industrial field due to its tedious annotation work, In this paper, we propose ADGAN, a unsupervised anomaly detection architecture using the variational autoencoder and the generative adversarial network which give great results in image generation task, and demonstrate whether the proposed network architecture identifies anomalous images well on MNIST benchmark dataset as well as our own welding defect dataset.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.14 no.6
• /
• pp.1408-1416
• /
• 1990

A powerful and efficient method for finding approximate solutions to initial-boundary-value problems in the transient coupled thermoelasticity is formulated in time domain using the finite element technique with time-marching strategy. The final system equations can be derived by the Guritin's variational principle using the definition of convolution integral. But, the finite element formulation for the equations of motion is modified by differentiating in time. Numerical results to some test problems are compared with analytical and other sophisticated approximate solutions. Stable responces are observed in all the given examples irrespective of incremental time steps and mesh shapes. In addition, it is shown that good numerical results are obtained even in coarser mesh or larger time step comparing to other numerical methods.

• Computational Structural Engineering
• /
• v.7 no.4
• /
• pp.103-111
• /
• 1994

Based on a variational principle of the consistent shear deformable discrete laminate theory derived in the companion paper Part I, a finite element procedure for the vibration analysis of laminated composite plates is presented. The present formulation takes the in-plane displacements of an arbitrary layer, the rotations of the cross section of each layer and transverse displacement of the plate as the state variables at a nodal point of finite element, resulting in total nodal degree of freedom of 2(n+l) +1 for the n-layered laminate. Thus, it allows to specify displacement boundary conditions of layer stretching and/or rotation of layer cross sections around the plate edge and/or lateral displacement. The developed procedure is applied to the free vibration problem for sandwich-type hybrid laminates composed of layers with drastically different material properties whose elasticity solutions are known. Comparison of analysis results with other FEM solutions showed that the present formulation yields better accuracy.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.9
• /
• pp.52-60
• /
• 2017

This study examined the structural stability of nonlocal magneto-electro-elastic nano plates on elastic foundations using first-order shear deformation theory. Navier's method has been used to solve the buckling loads for all edges simply supported boundary conditions. On the other hand, biaxial buckling analysis of nano-plates has beenrarely studied. According to the Maxwell equation and the magneto-electro boundary condition, the change inthe magnetic and electric potential along the thickness direction of the magneto-electro-elastic nano plate wasdetermined. To reformulate the elasticity theory of the magneto- electro-elastic nano plate, the differential constitutive equation of Eringen was used and the governing equation of the nonlocal elasticity theory was studied using variational theory. The effects of the elastic foundation arebased on Pasternak's assumption. The relationship between nonlocal theory and local theory was analyzed through calculation results. In addition, structural stability problems were investigated according to the electric and magnetic potentials, nonlocal parameters, elastic foundation parameters, and side-to-thickness ratio. The results of the analysis revealedthe effects of the magnetic and electric potential. These calculations can be used to compare future research on new material structures made of magneto-electro-elastic materials.

• Economic and Environmental Geology
• /
• v.54 no.1
• /
• pp.77-89
• /
• 2021

In this study, a method has been proposed to improve the performance of hydraulic property estimation model developed by Jeong et al. (2020). In their study, low-dimensional features of the annual groundwater level (GWL) fluctuation patterns extracted based on a Denoising autoencoder (DAE) was used to develop a regression model for predicting hydraulic properties of an aquifer. However, low-dimensional features of the DAE are highly dependent on the precipitation pattern even if the GWL is monitored at the same location, causing uncertainty in hydraulic property estimation of the regression model. To solve the above problem, a process for generating the GWL fluctuation pattern for conditioning the precipitation is proposed based on a conditional variational autoencoder (CVAE). The CVAE trains a statistical relationship between GWL fluctuation and precipitation pattern. The actual GWL and precipitation data monitored on a total of 71 monitoring stations over 10 years in South Korea was applied to validate the effect of using CVAE. As a result, the trained CVAE model reasonably generated GWL fluctuation pattern with the conditioning of various precipitation patterns for all the monitoring locations. Based on the trained CVAE model, the low-dimensional features of the GWL fluctuation pattern without interference of different precipitation patterns were extracted for all monitoring stations, and they were compared to the features extracted based on the DAE. Consequently, it can be confirmed that the statistical consistency of the features extracted using CVAE is improved compared to DAE. Thus, we conclude that the proposed method may be useful in extracting a more accurate feature of GWL fluctuation pattern affected solely by hydraulic characteristics of the aquifer, which would be followed by the improved performance of the previously developed regression model.