• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.30 no.8
• /
• pp.30-36
• /
• 2002

Wind tunnel model has some difference in shape compared to the real flight vehicle because of model support system for testing. The support system can make some differences in the measured forces and moments to the flight test data. There are several correction methods involved such as cavity pressure correction and model support interference. Internal balance and belly sting support were used for this wind tunnel test and three types of model support correction methods, variable sting thickness method, dummy sting method, and wire support method, were compared. Variable sting thickness method is well matched with wire support method, which is known for almost interference free.

• Communications for Statistical Applications and Methods
• /
• v.10 no.3
• /
• pp.873-878
• /
• 2003

LS-SVM(Least-Squares Support Vector Machine) has been used as a promising method for regression as well as classification. Suykens et al.(2000) used only the magnitude of residuals to obtain SVs(Support Vectors). Suykens' method behaves well for homogeneous model. But in a heteroscedastic model, the method shows a poor behavior. The present paper proposes a new method to get SVs. The proposed method uses the variance of noise as well as the magnitude of residuals to obtain support vectors. Through the simulation study we justified excellence of our proposed method.

• Structural Engineering and Mechanics
• /
• v.88 no.2
• /
• pp.189-200
• /
• 2023

Silicate glass is usually a brittle and plate-like material, and it is difficult to measure the elastic modulus by the traditional method. This paper develops a test method for the glass elastic modulus based on the fundamental frequency of the cantilever beam with an elastic support and a free end. The method installs the beam-type specimen on a semi-rigid support to form an elastic support-free end beam. The analytic solution of the stiffness coefficients of the elastic support is developed by the fundamental frequency of the two specimens with known elastic modulus. Then, the glass elastic modulus is measured by the fundamental frequency of the specimens. The method significantly improves the measurement accuracy and is suitable for the elastic modulus with the beam-type specimen whether the glass is homogeneous or not. Several tests on the elastic modulus measurement are conducted to demonstrate the reliability and validity of the test method.

• Korean Journal Plant Pathology
• /
• v.14 no.6
• /
• pp.676-681
• /
• 1998

Disease incidence of Codonopsis lanceolata was surveyed at the major cultivating fields in Chonbuk province in 1996 to 1997. The main diseases of Codonopsis lanceolata were ovserved as leaf spot caused by Septoria codonopsis, anthracnose by Glomerella cingulata, brown leaf spot by Cercospora sp., rust by Coleosporium koreanum, powdery mildew by Erysiphe sp., Fusarium wilt caused by Fusarium oxyporum, and white root rot by Sclerotium rolfsii. Anthracnose, leaf spot and brown leaf spot occurred severely on leaves from early July to late August. They were caused early fallen leaves. Fusarium wilt and white root rot occurred severely on stem and below the soil line in late August. They resulted in withering to death or chlorosis and fallen of leaves. Disease incidence of Codonopsis lanceolata was also substantially different in occurrence with a method of cultivation in late growth stage. Fusarium wilt and white root rot were more severe with a method of no support cultivation than those with a method of support cultivation with a stick. Fusarium wilt occurred 48.8% in a method of no support cultivation but 3.1% in a method of support cultivation with a stick. And white root rot occurred 18.9% in a method of no support cultivation but 0.3% in a method of no support cultivation with a stick. Thus, it proved that soil-borne diseases could be controlled support cultivation with a stick.

• Proceedings of the IEEK Conference
• /
• 2002.06c
• /
• pp.135-138
• /
• 2002

This paper addresses incremental learning method for regression. SVM(support vector machine) is a recently proposed learning method. In general training a support vector machine requires solving a QP (quadratic programing) problem. For very large dataset or incremental dataset, solving QP problems may be inconvenient. So this paper presents an incremental support vector learning method for function approximation problems.

• Journal of IKEEE
• /
• v.11 no.4
• /
• pp.227-234
• /
• 2007

In this paper, to support a QoS level appropriate to the user in Wireless Home Network based Wireless LAN, we propose a QoS support architecture which includes Wired Network and Wireless Network. Actually, an important problem to support QoS in Wireless Home Network is approached not only on a MAC level in Wireless LAN but also on a integrated method to combine Network layer with Datalink layer. By applying the integrated QoS support method, it is possible to provide QoS support architecture using a Wireless LAN terminal with a minimum changing, and the proposed scheme has advantage of QoS support method, which is more superior than a existing scheme to support QoS in MAC level of Wireless LAN. Simulations results show that overall performance of the proposed scheme can be improved.

• Journal of the Korean Geotechnical Society
• /
• v.28 no.12
• /
• pp.99-110
• /
• 2012

Inclined Earth Retaining Structure Method (IER method, briefly) is developed in order to improve the existing earth retaining method. In IER method, there are three main structures, front support, back support, and head binding. Especially, back support acts the role that reduces the earth pressure acting on the front support. In this study, the stability according to the installation angle and stiffness of front or back support is analysed by model tests. By the test results, it is known that inclined back support is very effective to reduce the earth pressure acting on the front support. Especially, the effect of the stiffness and installation angle of back support is analysed.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.8 no.3
• /
• pp.196-201
• /
• 2008

Statistical learning theory has three analytical tools which are support vector machine, support vector regression, and support vector clustering for classification, regression, and clustering respectively. In general, their performances are good because they are constructed by convex optimization. But, there are some problems in the methods. One of the problems is the subjective determination of the parameters for kernel function and regularization by the arts of researchers. Also, the results of the learning machines are depended on the selected parameters. In this paper, we propose an efficient method for objective determination of the parameters of support vector clustering which is the clustering method of statistical learning theory. Using evolutionary algorithm and bootstrap method, we select the parameters of kernel function and regularization constant objectively. To verify improved performances of proposed research, we compare our method with established learning algorithms using the data sets form ucr machine learning repository and synthetic data.

• Geomechanics and Engineering
• /
• v.22 no.1
• /
• pp.35-48
• /
• 2020

This paper provides methods for the deterministic and reliability-based design of the support pressures necessary to prevent tunnel face collapse. The deterministic method is developed by extending the use of the unique load multiplier, which is embedded within OptumG2/G3 with the intention of determining the maximum load that can be supported by a system. Both two-dimensional and three-dimensional examples are presented to illustrate the applications. The obtained solutions are validated according to those derived from the existing methods. The reliability-based method is developed by incorporating the Response Surface Method and the advanced first-order second-moment reliability method into the bisection algorithm, which continuously updates the support pressure within previously determined brackets until the difference between the computed reliability index and the user-defined value is less than a specified tolerance. Two-dimensional reliability-based support pressure is compared and validated via Monte Carlo simulations, whereas the three-dimensional solution is compared with the relationship between the support pressure and the resulting reliability index provided in the existing literature. Finally, a parametric study is carried out to investigate the influences of factors on the required support pressure.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.51 no.12
• /
• pp.630-637
• /
• 2002

This paper proposed a new diffusion model considering DSM rebate program's support conditions. The proposed method used some aspects of the rebate program such as support qualifications, annual support volume, and support level per appliance as following : The support qualifications were limited as the consumer which can get the rebate program's benefit, the annual support volume was constrained as the fixing budget and the support level per appliance was considered by high-efficient appliance actuality price. This paper also proposed a new method that used neural network as its parameter estimation moth[,4 for the diffusion model. The diffusion model and its parameter estimation method are expected to be able to analyze the diffusion characteristics of high-efficient appliance through the rebate program and the effects of rebate program's support conditions. Also, these will be able to evaluate the impacts and to analyze the cost-effectiveness of Energy Efficiency Demand-Side Management(EEDSM) resources. The case study is performed on the high-efficient lighting appliance rebate program of Korea by using the suggested diffusion model and estimation method and thus verified its validity.