• Journal of the Korean Geotechnical Society
• /
• v.17 no.1
• /
• pp.47-56
• /
• 2001

본 논문에서는 단일 및 군말뚝의 수평변위를 예측하기 위하여 신경망 학습속도의 향상과 지역 최소점 수렴을 방지하는 Readjusting 기법을 적용한 인공신경망을 도입하였다. 이 인공신경망을 M-EBPNN 이라고 한다. M-EBPNN에 의한 결과는 낙동강 모래지반에서 단일 및 군말뚝에 대하여 수행한 일련의 모형실험결과와 비교하였으며, 그리고 신경망의 학습속도와 지역 최소점의 수렴성을 평가하기 위하여 오류 역전파 신경망(EBPNN)의 결과와도 비교 분석하였다. M-EBPNN의 적용성 검증을 위하여 200개의 모형실험결과들을 이용하였으며, 신경망의 구조는 EBPNN의 구조와 동일한 한 개의 입력층과 두 개의 은닉층 그리고 한 개의 출력층으로 구성되었다. 전체 데이터의 25%, 50% 그리고 75% 결과는 각각 신경망의 학습에 이용되었으며 학습에 이용하지 않은 데이터들은 예측에 이용되었다. 그리고, 신경망의 최적학습을 위하여 적합한 은닉층의 뉴런 수와 학습률은 EBPNN에서 결정한 값들을 본 신경망에 이용하였다. 해석결과들에 의하면, 동일한 학습패턴에서의 M-EBPNN이 학습 반복횟수는 EBPNN 보다 최고 88% 감소하였으며 지역 최소점에 수렴하는 현상은 거의 나타나지 않았다. 따라서, 인공신경망 모델이 수평하중을 받는 말뚝의 수평변위 예측에 적용될 수 있는 가능성을 보여 주었다.

• The Journal of Engineering Geology
• /
• v.17 no.2 s.52
• /
• pp.197-204
• /
• 2007

Tunnel deformation happens by excavation. After installation of support, tunnel is gradually stabilized over time. Effect of excavation on tunnel behavior decreases as increase of distance from face. If the time that the displacement converges by tunnel stabilization is estimated, processes after stabilization can be advanced and economic loss can be reduced. In this study, the distance of displacement convergent point from face in the tunnel constructed on sedimentary rock is estimated using control chart method. As the results of analysis using a control of chart, displacements in a sedimentary rock tunnel are converged within 100 m from each tunnel face.

• Journal of the Korean GEO-environmental Society
• /
• v.21 no.5
• /
• pp.13-21
• /
• 2020

In this study, the reliability of the lateral spring constant (k1) applied during design of pile foundation for bridge abutment was evaluated. To do this, the reliability of the factors related to the prediction of the lateral displacement of the abutment pile foundation, which was designed based on the displacement method proposed by Chang (1937), was analyzed. The data used for analysis were the design statements of ◯◯ bridge and ◯◯ IC2 bridge. Then, it was derived by comparing with the numerical analysis (p-y analysis) based on the basic data.

• Tunnel and Underground Space
• /
• v.12 no.2
• /
• pp.99-106
• /
• 2002

To use the early measured data of tunnel deformation in but analysis, the relationship between these values find final deformation data were studied. Panet\`s exponential and fraction equations successfully approximate the convergence of the hard rock tunnels. Measured deformation data of ID location, $U_{1D}$ show that they can be lilted to linear equations but should not be used to estimate potential deformation before measurement, $C_{0}$. Early measured data $U_{1D}$ $U_{2D}$ , and final deformation $ U_{L}$ showed linear correlations. It proved that estimated data of final deformation from early measured ones can be used as input parameters for back analysis.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.141-143
• /
• 2021

본 논문에서는 목조 문화재의 변위량을 감지할 수 있는 앙상블 딥러닝 모델 모델을 제안한다. 우선 총 2개의 서로 다른 사전 학습된 합성 곱 신경망을 사용하여 입력 영상에 대한 심층 특징들을 추출한다. 그 이후 2개의 서로 다른 심층 특징들을 결합하여 하나의 특징 벡터를 생성한다. 그 이후 합쳐진 특징 벡터는 완전 연결 계층의 입력 값으로 들어와서 최종적으로 변위의 심각 단계에 대한 예측을 수행하게 된다. 데이터 셋으로는 충주시 근처의 문화재에 방문해서 수집한 목조 문화재 이미지를 가지고 정상 및 비정상으로 구분한 데이터 셋을 사용하였다. 실험 결과 앙상블 딥러닝 기법을 사용한 모델이 앙상블 기법을 사용하지 않는 모델보다 더 좋은 성능을 나타냄을 확인하였다. 이러한 결과로부터 우리가 제안한 방법이 목재 문화재의 변위량 예측에 있어서 매우 적합함을 보여준다.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2001.11b
• /
• pp.1384-1387
• /
• 2001

Presented is a method to predict strain responses from displacement measurements on a mechanical structure. The method consists of forming a transformation matrix, which is calculated from displacement and strain modal matrices. The modal matrices can be obtained by either finite element analysis or modal testing. One disadvantage of the method is that it requires displacements on all measuring points be measured simultaneously. The strain prediction method is applied to a simple simulated system.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.12 no.2
• /
• pp.69-80
• /
• 2008

In this study, improved capacity spectrum method (CSM) is proposed. The method can account for higher mode contribution to the seismic response of MDOF systems. The CSM has been conveniently used for determining maximum roof displacement using both demand spectrum and capacity curve of equivalent SDOF system. Unlike the conventional CSM, the maximum roof displacement is determined without iteration using inelastic displacement ratio and R factor calculated from demand spectrum and capacity curve. Three moment resisting steel frames of 3-, 9- and 20-stories are considered to test the accuracy of the proposed method. Nonlinear response history analysis (NL-RHA) for three frames is also conducted, which is considered as an exact solution. SAC LA 10/50 and 2/50 sets of ground motions are used. Moreover, this study estimates maximum story drift ratios (IDR) using ATC-40 CSM and N2-method and compared with those from the proposed method and NL-RHA. It shows that the proposed CSM estimates the maximum IDR accurately better than the previous methods.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.8 no.2
• /
• pp.101-113
• /
• 2006

Arching effect occurs around the unsupported excavation surface near to tunnel face when a tunnel is excavated in a stable rock mass. If a weak fracture zone exists in front of tunnel face, a displacement occurs between tunnel face and weak fracture zone due to stress concentration. If three-dimensional absolute coordinates (longitudinal, transverse, vertical direction) is measured at tunnel face by geodetic method, the ground change in front of the tunnel face can be predicted by analysing three-dimensional absolute displacement. The purpose of this study is to verify the analysis method of three-dimensional absolute displacement by comparing the trend of displacement ratio at crown and sidewall of tunnel and the influence line/trend line of crown settlement compared with TSP results in the same section.

• Journal of the Korean Geotechnical Society
• /
• v.21 no.3
• /
• pp.99-108
• /
• 2005

One of the most important and difficult tasks in designing underground structure is the estimation of engineering properties of the ground. The main purpose of this study is to propose a new back-analysis technique in tunnelling to estimate geotechnical parameters around a tunnel. In this study, the Extended Bayesian Method, which appropriately combines objective information with subjective one, is adopted to optimize engineering parameters. By using only relative convergence data measured during tunnelling as input values in back-analysis, inevitable errors in absolute convergence estimation are excluded and 3-dimensional numerical analysis is applied to consider a trend of relative convergence occurrence. Finally, 3-dimensional back-analysis technique using relative convergence is proposed and evaluated using a hypothetical site.

• The Journal of Engineering Geology
• /
• v.22 no.1
• /
• pp.39-48
• /
• 2012

The behaviors of a diaphragm wall and a contiguous pile wall such as CIP(Case-in-place pile) and SCW(Soil-cement wall), applied to the top-down construction method, were analyzed using the SUNEX program, which is widely used to design earth retaining walls. Four types of earth pressures, as described by Rankine (1857), Terzaghi and Peck (1967), Tchbotarioff (1973), and Hong and Yun (1995a), were applied to the analysis program to predict the lateral displacement of walls. The results show that the displacements of an earth retaining walls vary with the applied earth pressure. The predicted lateral displacement based on Hong & Yun's (1995a) earth pressure is similar to the measured displacement. Therefore, the actual lateral displacement of an earth retaining wall, as applied to top-down construction method, can be accurately predicted by using an analysis program considering Hong and Yun's (1995a) earth pressure.