• Journal of Korean Association for Spatial Structures
• /
• v.20 no.2
• /
• pp.51-58
• /
• 2020

Recently, deep learning that is the most popular and effective class of machine learning algorithms is widely applied to various industrial areas. A number of research on various topics about structural engineering was performed by using artificial neural networks, such as structural design optimization, vibration control and system identification etc. When nonlinear semi-active structural control devices are applied to building structure, a lot of computational effort is required to predict dynamic structural responses of finite element method (FEM) model for development of control algorithm. To solve this problem, an artificial neural network model was developed in this study. Among various deep learning algorithms, a recurrent neural network (RNN) was used to make the time history response prediction model. An RNN can retain state from one iteration to the next by using its own output as input for the next step. An eleven-story building structure with semi-active tuned mass damper (TMD) was used as an example structure. The semi-active TMD was composed of magnetorheological damper. Five historical earthquakes and five artificial ground motions were used as ground excitations for training of an RNN model. Another artificial ground motion that was not used for training was used for verification of the developed RNN model. Parametric studies on various hyper-parameters including number of hidden layers, sequence length, number of LSTM cells, etc. After appropriate training iteration of the RNN model with proper hyper-parameters, the RNN model for prediction of seismic responses of the building structure with semi-active TMD was developed. The developed RNN model can effectively provide very accurate seismic responses compared to the FEM model.

• Journal of Soil and Groundwater Environment
• /
• v.20 no.7
• /
• pp.112-120
• /
• 2015

Statistical trend analysis using the data from the National Groundwater Quality Monitoring Network (NGQMN) of Korea was conducted to establish a new groundwater quality management framework. Sen’s test, a non-parametric statistical method for trend analysis, was used to determine the linear trend of the groundwater quality data. The analysis was conducted at different confidence levels (i.e., at 70, 80, 90, 95, and 99% confidence levels) for three of groundwater quality parameters, i.e., nitrate-nitrogen, chloride, and pH, which have sufficient time series of the NGQMN data between 2007 and 2013. The results showed that different trends can be determined for different depths even for the same monitoring site and the numbers of wells having significant trends vary with different confidence levels. The wells with increasing or decreasing trends were far less than the wells with no trend. Chloride had more wells with increasing trend than other parameters. On the other hand, nitrate-nitrogen had the most wells with increasing trend and concentration exceeding 75% of the threshold values (TVs). Based on the methodology used for this study, we suggest including groundwater TVs and trend analysis to evaluate groundwater quality and to establish an advanced groundwater quality management framework.

• Proceedings of the KIEE Conference
• /
• 2004.11c
• /
• pp.346-348
• /
• 2004

In this rarer, we introduce a new Fuzzy Polynomial Neural Networks (FPNNs)-like structure whose neuron is based on the Fuzzy Set-based Fuzzy Inference System (FS-FIS) and is different from that of FPNNs based on the Fuzzy relation-based Fuzzy Inference System (FR-FIS) and discuss the ability of the new FPNNs-like structurenamed Fuzzy Set-based Polynomial Neural Networks (FSPNN). The premise parts of their fuzzy rules are not identical, while the consequent parts of the both Networks (such as FPNN and FSPNN) are identical. This difference results from the angle of a viewpoint of partition of input space of system. In other word, from a point of view of FS-FIS, the input variables are mutually independent under input space of system, while from a viewpoint of FR-FIS they are related each other. In considering the structures of FPNN-like networks such as FPNN and FSPNN, they are almost similar. Therefore they have the same shortcomings as well as the same virtues on structural side. The proposed design procedure for networks' architecture involves the selection of appropriate nodes with specific local characteristics such as the number of input variables, the order of the polynomial that is constant, linear, quadratic, or modified quadratic functions being viewed as the consequent part of fuzzy rules, and a collection of the specific subset of input variables. On the parameter optimization phase, we adopt Information Granulation (IG) based on HCM clustering algorithm and a standard least square method-based learning. Through the consecutive process of such structural and parametric optimization, an optimized and flexible fuzzy neural network is generated in a dynamic fashion. To evaluate the performance of the genetically optimized FSPNN (gFSPNN), the model is experimented with using gas furnace process dataset.

• Composites Research
• /
• v.24 no.4
• /
• pp.17-22
• /
• 2011

In this paper, as a basic research to apply the composite sandwich to underwater vehicle, the manufacturing, analysis and test methods, and weight saving effect of a composite sandwich cylinder under external pressure were studied. A two-step manufacturing method to prevent the wrinkling of the sandwich cylinder face was proposed and the three cylinders were made and tested. Finite element results based on the shell and solid model using MSC.Nastran were compared with test results. The comparison showed that the linear finite element analysis using the shell and solid elements can predict the buckling pressure of the sandwich cylinder with approximately 3% difference. The parametric study of the filament wound cylinders revealed that the composite sandwich can reduce the weight of the cylinder more than 30% compared with the filament wound cylinder supporting the same pressure.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.26 no.11
• /
• pp.2305-2311
• /
• 2002

The pressure tube is a major component of the CANDU reactor, which supports nuclear fuel bundle and heavy water coolant. Pressure tubes are installed horizontally inside the reactor and only selected samples are periodically examined during in-service inspection. In this respect, a probabilistic safety assessment method is more appropriate fur the assessment of overall pressure tube safety. The failure behavior of CANDU pressure tubes, however, is governed by delayed hydride cracking which is the major difference from pipings and reactor pressure vessels. Since the delayed hydride cracking has more widely distributed governing parameters, it is impossible to apply a general PFM methodology directly. In this paper, a PFM methodology for the safety assessment of CANDU pressure tubes is introduced by applying Monte Carlo simulation in determining failure probability Initial hydrogen concentration, flaw shape and depth, axial and radial crack growth rate and fracture toughness were considered as probabilistic variables. Parametric study has been done under the base of pressure tube dimension and hydride precipitation temperature in calculating failure probability. Unstable fracture and plastic collapse are used for the failure assessment. The estimated failure probability showed about three-order difference with changing dimensions of pressure tube.

• Journal of the Korean Society for Library and Information Science
• /
• v.42 no.2
• /
• pp.371-392
• /
• 2008

The books and circulation-related data in the Online Public Access Catalog System of the H-Public library were collected and analyzed in this paper. The methods which may be applied to the Customer Relationship Management in a public library based on the results were also suggested here. Collected data were 57,927 bibliographic data of books, 11,871 patron data and 27,145 circulation records. The type of collections. circulation frequencies, total number of circulated books, publication year, and use factor as relation factor from the analyzed data of circulation records were extracted. They were also analyzed, and verified by various statistical methods such as correlation coefficient, non-parametric method, etc.

• The Korean Journal of Applied Statistics
• /
• v.31 no.1
• /
• pp.41-65
• /
• 2018

In Korea, as the mortality rate improves in a shorter period of time than in developed countries, it is important to consider the selection of the time series as well as the model selection in the mortality projection. Therefore, this study proposed a method using the multiple regression model in respect to the selection of the time series period. In addition, we investigate the problems that arise when various time series are used based on the Lee-Carter (LC) model, the kinds of LC model along with Lee-Miller (LM) and Booth-Maindonald-Smith (BMS), and the non-parametric model such as functional data model (FDM) and Coherent FDM, and examine differences in the age-specific mortality rate and life expectancy projection. Based on the analysis results, the age-specific mortality rate and predicted life expectancy of men and women are calculated for the year 2030 for each model. We also compare the mortality rate and life expectancy of the next generation provided by Korean Statistical Information Service (KOSIS).

• The KIPS Transactions:PartD
• /
• v.8D no.4
• /
• pp.393-400
• /
• 2001

This paper presents a stochastic model for the software failure phenomenon based on a nonhomogeneous Poisson process (NHPP). The failure process is analyzed to develop a suitable mean value function for the NHPP ; expressions are given for several performance measure. Actual software failure data are compared with generalized model by Goel dependent on the constant reflecting the quality of testing. The performance measures and parametric inferences of the new models, Rayleigh and Gumbel distributions, are discussed. The results of the new models are applied to real software failure data and compared with Goel-Okumoto and Yamada, Ohba and Osaki models. Tools of parameter inference was used method of the maximun likelihood estimate and the bisection algorithm for the computing nonlinear root. In this paper, using the sum of the squared errors, model selection was employed. The numerical example by NTDS data was illustrated.

• Journal of Korean Society of Steel Construction
• /
• v.9 no.2 s.31
• /
• pp.181-191
• /
• 1997

Before cables are constructed, tower of suspension bridge is behaved as a cantilever type. Buffeting occured by unsteady loading of the tower due to velocity fluctuation in the oncoming flow has a wind velocity consistent with fundamental frequency of the tower and may give rise to large response by the tower resonance. To reduce the dynamic response by buffeting, the behavior of tower with TMD(Tuned Mass Damper) has studied using finite element method in time domain. The buffeting was obtained by transforming the velocity spectrum in frequency domain to random variable in certain time domain. The most probable maximum displacement which can be occured during the time interval was obtained using peak factor. The optimum location for TMD installation and TMD specification were decided by parametric study. Also, the effect of vibration control about various wind velocity was studied by the TMD which has optimum specification and location.

• Earthquakes and Structures
• /
• v.9 no.4
• /
• pp.831-849
• /
• 2015

Seismic-design procedures for walls require that the confinement in the critical (plastic hinge) regions should extend over a length in the compression zone of the cross section at the wall base where concrete strains in the Ultimate Limit State (ULS) exceed the limit of 0.0035. In a performance-based framework, confinement is linked to required curvature ductility so that the drift demand at the performance point of the structure for the design earthquake may be met. However, performance of flexural walls in the recent earthquakes in Chile (2010) and Christchurch (2011) indicates that the actual compression strains in the critical regions of many structural walls were higher than estimated, being responsible for several of the reported failures by toe crushing. In this study, the method of estimating the confined region and magnitude of compression strain demands in slender walls are revisited. The objective is to account for a newly identified kinematic interaction between the normal strains that arise in the compression zone, and the lumped rotations that occur at the other end of the wall base due to penetration of bar tension yielding into the supporting anchorage. Design charts estimating the amount of yield penetration in terms of the resulting lumped rotation at the wall base are used to quantify the increased demands for compression strain in the critical section. The estimated strain increase may exceed by more than 30% the base value estimated from the existing design expressions, which explains the frequently reported occurrence of toe crushing even in well confined slender walls under high drift demands. Example cases are included in the presentation to illustrate the behavioral parametric trends and implications in seismic design of walls.