• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.18 no.2
• /
• pp.201-211
• /
• 2005

Due to the complex mechanism and various parameters that affect shear behavior of reinforced concrete (RC) members, models on shear tend to be complex and difficult to utilize for design of structural members, and empirical relationships formulated with limited test data often work lot members having a specific range of influencing parameters on shear. As an alternative approach tot solving this problem, artificial neural networks have been suggested by some researchers. In this paper, artificial neural networks were used to predict shear strengths of RC beams without shear reinforcement. Especially, a large database that consists of shear test results of 398 RC members without shear reinforcement was used for artificial neural network analysis. Three well known approaches for shear strength of RC members, ACI 318-02 shear provision, Zsutiy's equation, and Okamura's relationship, are also evaluated with test results in the shear database and compared with neural network approach. While ACI 318-02 provided inaccurate predictions for RC members without shear reinforcement, the empirical equations by Zsutty and Okamura provided more improved prediction of Shear strength than ACI 318-02. The artificial neural networks, however provided the best prediction of shear strengths of RC beams without shear reinforcement that was closest to test results.

• The Korean Journal of Applied Statistics
• /
• v.22 no.2
• /
• pp.415-424
• /
• 2009

As one of the unsupervised learning neural network methods, self-organizing maps(SOM) are applied to various fields. It reduces the dimension of multidimensional data by representing observations on the low dimensional manifold. On the other hand, the minimal spanning tree(MST) of a graph that achieves the most economic subset of edges connecting all components by a single open loop. In this study, we apply the MST technique to SOM with subnodes. We propose SOM's with embedded MST and a distance measure for optimum choice of the size and shape of the map. We demonstrate the method with Fisher's Iris data and a real gene expression data. Simulated data sets are also analyzed to check the validity of the proposed method.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.570-573
• /
• 2022

This paper presents the identification of void diameters for a cast-resin transformer using an artificial neural network (ANN) model. A PD signal was measured by the Rogowski coil sensor which has the planar and thin structures fabricated on a printed circuit board (PCB), and the PD electrode system was fabricated to simulate a PD defect by a void. In addition, void samples with different diameters were fabricated by injecting air in a cylindrical aluminum frame using a syringe during the epoxy curing process. To identify the diameter of void defects, PD characteristics such as the discharge magnitude, pulse count, and phase angle were extracted and back propagation algorithm (BPA) was designed using virtual instrument (VI) based on the Labview program. From the experimental results, the BPA algorithm proposed in this paper has over 90% accurate rate to identify the diameter of void defects and is expected to use reference data of maintenance and replacement of insulation for cast-resin transformers in the on-site PD measurement.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.29 no.6
• /
• pp.667-675
• /
• 2011

In this paper, we propose an automatic image-to-image registration between high resolution multi-sensor images. To do this, TerraSAR-X image was shifted according to the initial translation differences of the x and y directions between images estimated using Mutual Information method. After that, the Canny edge operator was applied to both images to extract linear features. These features were used to design a cost function that finds matching points based on the similarities of their locations and gradient orientations. For extracting large number of evenly distributed matching points, only one point within each regular grid constructed throughout the image was extracted to the final matching point pair. The model, which combined the piecewise linear function with the global affine transformation, was applied to increase the accuracy of the geometric correction, and the proposed method showed RMSE lower than 5m in all study sites.

• Journal of the Korean earth science society
• /
• v.23 no.8
• /
• pp.683-696
• /
• 2002

In this work. we investigated the ozone data sets that exceeded ambient air quality standards from 31 air quality monitoring stations dispersed across the Seoul metropolitan city during the period covering 1990 and 2000. To specifically describe spatial dependency of high level O$_3$ occurrence, we grouped our data into four different geographical ozone exceedance is much longer in SW than the other three sectors. When we compared the exceedance data in terms of occurrence frequency, the month of maximum frequency differed slightly among different sectors. Examination of long-term exceedance trend indicated that its frequency increased continuously from all sectors over the past years, although slightly opposite patterns existed in their absolute values. Most importantly, its peak occurrence frequency seemed to center in very recent years such as 1998 (NE sector) and 2000 (ail pattern sectors except NE). Consequently, we were able to describe the existence of certain patterns of ozone exceedance data sets in terms of both temporal and spatial scales.

• Journal of the Korean earth science society
• /
• v.24 no.3
• /
• pp.190-204
• /
• 2003

The PM (particulate matter) concentration data sets exceeding the Daily Air Quality Guidance Levels (i.e., established by the Korean Ministry of Environment) were selected from 31 air qality monitoring stations in Seoul from 1990 to 2000. (For reference, the 24hr environmental standard values of PM$_{10}$ and TSP are 150 and 300${\mu}$g/m$^{3}$, respectively.) When the data sets were compared between land use types, both PM fractions were exceeded most frequently in residential areas. However, the highest TSP concentration was measured at industrial areas (351.0${\pm}$35.9${\mu}$g m$^{3}$), while the highest PM$_{10}$ concentration was measured in residential areas (182.9${\pm}$42.4${\mu}$g m$^{3}$). When the temporal distribution patterns of the exceedance data sets were compared to those measured routinely (without any discrimination based on exceedance criteria), large differences were present. It was demonstrated that the occurrences of exceedance data sets increased rather significantly in recent years.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2000.04a
• /
• pp.680-685
• /
• 2000

최근 그 심각성이 커지고 있는 해킹피해를 줄이기 위한 한 방법으로 시스템에 침입한 불법적 사용을 탐지하는 연구가 활발히 진행되고 있다. 침입을 탐지하는 방법으로는 오용탐지와 비정상행위 탐지가 있는데 비정상행위 탐지를 위해서는 정보수집의 정확성, 신속성과 함께 다량의 정보들로부터 필요한 정보를 추출하고 축약하는 것이 중요하다. 본 논문에서는 감사기록 도구인 BSM으로부터 정보를 추출하고 자기조직화 신경망을 이용하여 다차원의 정보를 저차원정보로 축약.변환하는 방법에 대한 실험적인 검증을 시도하였다. 또한 BSM에서 얻을 수 있는 데이터의 유용성을 조사하기 위하여 축약된 감사자료에 의한 탐지성능을 살펴보았다. 실험결과, 시스템 호출 및 파일관련 정보의 축약이 탐지성능향상에 크게 기여하는 중요한 척도임을 알 수 있었으며 각 척도마다 탐지성능이 좋은 맵의 크기가 다름을 알 수 있었다. 이러한 축약된 정보는 여러 정상행위 모델링방법에 의해 유용하게 사용될 수 있을 것이다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2005.05b
• /
• pp.835-839
• /
• 2005

일반적으로 분포형 모형에서는 유역을 공간적으로 작은 계산단위로 분할하고, 각 단위에 대해 모형의 이론식을 전개하여 풀이하게 된다. 이 계산단위는 일반적으로 입력자료의 공간적인 해상도보다 크기 때문에 어느 정도 수준까지의 취합을 내포하게 된다. SWAT에서도 수문응답단위 (HRU; Hydrological Response Unit)라는 계산 단위를 통하여 모형 입력 매개변수를 생성하고, 모의를 수행한다. 따라서, 본 연구에서는 SWAT 모형의 거동 특성과 유역별 적정한 수준의 소유역 분할에 대한 기준을 제시할 목적으로 경안천 유역과 보청천 유역을 선정하여, 각 유역변 소유역 수 및 HRU 수에 따른 연평균 유출과 유사의 변화를 검토하였다. 검토 결과, 경안천 유역에 대해서는 SWAT 모형의 적용시에 하천망 생성을 위한 임계면적을 300 ha 이하로 두어 55개 이상의 소유역으로 분할하고, HRU 생성을 위한 토지이용과 토양 면적비는 $8\%$ 이하로 설정하여야 안정적인 유출과 유사 모의가 가능하며, 보청천에 대해서는 임계면적을 5,000 ha 이하로 하여 최소 5개 이상의 소유역으로 분할하고, 토지이용과 토양 면적비는 $1\%$ 이하로 설정하여 HRU를 생성함으로써 안정적인 유출 및 유사 모의가 가능한 것으로 나타났다. 이상의 결과와 같이 적절한 수준의 소유역 분할과 HRU 생성에 대한 기준을 제시함으로써, 모형의 모의 결과의 신뢰도를 크게 감소시키지 않으면서 모형의 입력자료 구축시간과 모형 구동시간을 단축함으로써 모형의 적용 효율을 높일 수 있을 것으로 판단된다.

• Journal of Korea Water Resources Association
• /
• v.36 no.2
• /
• pp.161-172
• /
• 2003

This study has evaluated the stream gauge network with the main emphasis on if the current stream gauge network can catch the runoff characteristics of the basin. As the evaluation of the stream gauge network in this study does not consider a special purpose of a stream gauge, nor the effect from a hydraulic structure, it becomes an optimization of current stream gauge network under the condition that each stream gauge measures the natural runoff volume. This study has been applied to the Nam-Han River Basin for the optimization of total 31 stream gauge stations using the entropy concept. Summarizing the results are as follows. (1) The unit hydrograph representing the basin response from rainfall can be transferred into a probability density function for the application of the entropy concept to optimize the stream gauge network. (2) Accurate derivation of unit hydrographs representing stream gauge sites was found the most important part for the evaluation of stream gauge network, which was assured in this research by comparing the measured and derived unit hydrographs. (3) The Nam-Han River Basin was found to need at least 28 stream gauge stations, which was derived by considering both the shape of the unit hydrograph and the runoff volume. If considering only the shape of the unit hydrograph, the number of stream gauges required decreases to 23.

• Journal of Korean Society of Transportation
• /
• v.17 no.4
• /
• pp.111-124
• /
• 1999

The Purpose of this thesis is to develop a vehicle classification algorithm using single Magnetometer detector during presence time of vehicle detection and is to examine a held application from field test. We collected data using Magnetometer detector on freeway and used digital data to change voltage values according to magnetic flux density in analysis. We collected these datum during the presence time and then obtained characteristics from wave form in these datum. Based on these characteristics, We used the following three methods for this a1gorithm :1. Template Matching Method,2. Neural Network Method using Back-propagation Algorithm 3. Complex Method using changed slope points and mixing method 1, 2. Of course, Before processing of over three methods, These data were processed normalizing by 20, 40 of size in only X axis and moving average by 0, 3, 4, 5 of size. Vehicle classification were Processed in three steps ; 2, 3, 5 types classification. In 2 types vehicle classification, recognition rate is 83% by template matching method.