• Title/Summary/Keyword: 요소망 의존성

Search Result 29, Processing Time 0.023 seconds

Finite Element Mesh Dependency in Nonlinear Earthquake Analysis of Concrete Dams (콘크리트 댐의 비선형 지진해석에서의 유한요소망 영향)

  • 이지호
    • Journal of the Korea Concrete Institute
    • /
    • v.13 no.6
    • /
    • pp.637-644
    • /
    • 2001
  • A regularization method based on the Duvaut-Lions viscoplastic scheme for plastic-damage and continuum damage models, which provides mesh-independent and well-posed solutions in nonlinear earthquake analysis of concrete dams, is presented. A plastic-damage model regularized using the proposed rate-dependent viscosity method and its original rate-independent version are used for the earthquake damage analysis of a concrete dam to analyze the effect of the regualarization and mesh. The computational analysis shows that the regularized plastic-damage model gives well-posed solutions regardless mesh size and arrangement, while the rate-independent counterpart produces mesh-dependent ill-posed results.

Topology Design Optimization of Nonlinear Thermo-elastic Structures (비선형 열탄성 연성구조의 위상 최적설계)

  • Moon, Min-Yeong;Jang, Hong-Lae;Kim, Min-Geun;Cho, Seon-Ho
    • Journal of the Computational Structural Engineering Institute of Korea
    • /
    • v.23 no.5
    • /
    • pp.535-541
    • /
    • 2010
  • In this paper, we have derived a continuum-based adjoint design sensitivity of general performance functionals with respect to Young' modulus and heat conduction coefficient for steady-state nonlinear thermoelastic problems. An adjoint equation for temperature and displacement fields is defined for the efficient computation of the coupled field design sensitivity. Through numerical examples, we investigated the mesh dependency of the topology optimization method in the thermoelastic problems. Also, comparing the dominant loading cases of thermal and mechanical ones, the loading dependency of topology design optimization in coupled multi-physics problems is investigated.

소프트웨어 공급망 보안 관리를 위한 기술 동향 조사와 향후 발전 방향 제언

  • Kang, WooSeong;Pang, Hyunkjun
    • Review of KIISC
    • /
    • v.32 no.5
    • /
    • pp.21-25
    • /
    • 2022
  • 디지털 트랜스포메이션(Digital Transformation)으로 인해 소프트웨어에 대한 의존성이 강화되면서 소프트웨어 공급망의 역할이 커지고 있다. 그러나 안전한 소프트웨어 개발 및 이용을 위한 소프트웨어 공급망 보안 관리는 사실상 어려운 실정이다. 공급망이 복잡해질수록 공급망 공격의 유형은 다각화되는 반면 공급망을 구성하는 공급업체 및 구성요소에 대한 식별 및 취약점 분석은 어려워지기 때문이다. 이에 저자는 국내외 소프트웨어 공급망 보안 관리를 위한 기술 동향에 대한 조사 분석을 수행하고 이를 기반으로 향후 적용할 수 있는 공급망 보안 관리 체계의 발전 방향에 대해 작성하였다.

Comparing the Performance of Artificial Neural Networks and Long Short-Term Memory Networks for Rainfall-runoff Analysis (인공신경망과 장단기메모리 모형의 유출량 모의 성능 분석)

  • Kim, JiHye;Kang, Moon Seong;Kim, Seok Hyeon
    • Proceedings of the Korea Water Resources Association Conference
    • /
    • 2019.05a
    • /
    • pp.320-320
    • /
    • 2019
  • 유역의 수문 자료를 정확하게 분석하는 것은 수리 구조물을 효율적으로 운영하기 위한 중요한 요소이다. 인공신경망(Artificial Neural Networks, ANNs) 모형은 입 출력 자료의 비선형적인 관계를 해석할 수 있는 모형으로 강우-유출 해석 등 수문 분야에 다양하게 적용되어 왔다. 이후 기존의 인공신경망 모형을 연속적인(sequential) 자료의 분석에 더 적합하도록 개선한 회귀신경망(Recurrent Neural Networks, RNNs) 모형과 회귀신경망 모형의 '장기 의존성 문제'를 개선한 장단기메모리(Long Short-Term Memory Networks, 이하 LSTM)가 차례로 제안되었다. LSTM은 최근에 주목받는 딥 러닝(Deep learning) 기법의 하나로 수문 자료와 같은 시계열 자료의 분석에 뛰어난 성능을 보일 것으로 예상되며, 수문 분야에서 이에 대한 적용성 평가가 요구되고 있다. 본 연구에서는 인공신경망 모형과 LSTM 모형으로 유출량을 모의하여 두 모형의 성능을 비교하고 향후 LSTM 모형의 활용 가능성을 검토하고자 하였다. 나주 수위관측소의 수위 자료와 인접한 기상관측소의 강우량 자료로 모형의 입 출력 자료를 구성하여 강우 사상에 대한 시간별 유출량을 모의하였다. 연구 결과, 1시간 후의 유출량에 대해서는 두 모형 모두 뛰어난 모의 능력을 보였으나, 선행 시간이 길어질수록 LSTM의 정확성은 유지되는 반면 인공신경망 모형의 정확성은 점차 떨어지는 것으로 나타났다. 앞으로의 연구에서 유역 내 다양한 수리 구조물에 의한 유 출입량을 추가로 고려한다면 LSTM 모형의 활용성을 보다 더 확장할 수 있을 것이다.

  • PDF

A Survey on Neural Networks Using Memory Component (메모리 요소를 활용한 신경망 연구 동향)

  • Lee, Jihwan;Park, Jinuk;Kim, Jaehyung;Kim, Jaein;Roh, Hongchan;Park, Sanghyun
    • KIPS Transactions on Software and Data Engineering
    • /
    • v.7 no.8
    • /
    • pp.307-324
    • /
    • 2018
  • Recently, recurrent neural networks have been attracting attention in solving prediction problem of sequential data through structure considering time dependency. However, as the time step of sequential data increases, the problem of the gradient vanishing is occurred. Long short-term memory models have been proposed to solve this problem, but there is a limit to storing a lot of data and preserving it for a long time. Therefore, research on memory-augmented neural network (MANN), which is a learning model using recurrent neural networks and memory elements, has been actively conducted. In this paper, we describe the structure and characteristics of MANN models that emerged as a hot topic in deep learning field and present the latest techniques and future research that utilize MANN.

Automated Structural Design System Using Fuzzy Theory and Neural Network (피지이론과 신경망을 이용한 구조설계의 자동화 시스템)

  • Lee, Joon-Seong
    • Journal of the Korean Society for Precision Engineering
    • /
    • v.15 no.12
    • /
    • pp.236-243
    • /
    • 1998
  • 퍼지 이론과 계산기하학적 수법에 의한 자동요소 생성법, 해석코드 및 상용 솔리드 모델러를 유기적으로 통합한 자동화된 설계시스템을 개발하였다. 본 시스템은 여러 가지 복합현상과 관련된 실제 구조물에 대한 설계기능을 갖고 있다. 정전장 해석, 변형해석 및 모드해석 등과 같은 해석하고자 하는 물리적인 현상에 의존한 형상모델이 자동적으로 유한요소모델로 변환되어 해석을 수행한다. 또한 신경망의 기능을 도입, 통합시킴으로써 설계해의 영역을 유용하게 제시하여 준다. 개발한 시스템은 정전 마이크로머쉰의 성능 평가에 적용하여 그 효용성을 검증하였다.

  • PDF

The DFN-DEM Approach Applied to Investigate the Effects of Stress on Mechanical and Hydraulic Rock Mass Properties at Forsmark, Sweden (암반균열망-개별요소법 수치실험을 통해 살펴본 스웨덴 포쉬마크지역 암반의 역학적 및 수리적 물성에 초기응력이 미치는 영향)

  • Min, K.B.;Stephansson, O.
    • Tunnel and Underground Space
    • /
    • v.21 no.2
    • /
    • pp.117-127
    • /
    • 2011
  • The purpose of this study is to demonstrate the effect of in-situ rock stresses on the deformability and permeability of fractured rocks. Geological data were taken from the site investigation at Forsmark, Sweden, conducted by Swedish Nuclear Fuel and Waste Man-agement Company (SKB). A set of numerical experiments was conducted to determine the equivalent mechanical properties (essentially, elastic moduli and Poisson's ratio) and permeability, using a Discrete Fracture Network-Discrete Element Method (DFN-DEM) approach. The results show that both mechanical properties and permeability are highly dependent on stress because of the hyperbolic nature of the stiffness of fractures, different closure behavior of fractures, and change of fluid pathways caused by deformation. This study shows that proper characterization and consideration of in-situ stress are important not only for boundary conditions of a selected site but also for the understanding of the mechanical and hydraulic behavior of fractured rocks.

Performance Evaluation of DiffServ Networks Considering Self-Similar Traffic Characteristics (자기유사 트래픽 특성을 고려한 차등서비스 망의 성능 평가)

  • Park, Jeong-Sook;Jeon, Yong-Hee
    • The Journal of Korean Institute of Communications and Information Sciences
    • /
    • v.33 no.5B
    • /
    • pp.344-355
    • /
    • 2008
  • In this paper, we are dealing with the problems of performance evaluation of Differentiated Services(DiffServ) networks. For successful performance evaluation, the ability to accurately represent "real" traffic on the network by suitable traffic models is an essential ingredient. Many research results on the nature of real traffic measurements demonstrated LRD(long-range dependence) property for the Internet traffic including Web, TELNET, and P2P traffic. The LRD can be effectively represented by self-similarity. In this paper, we design and implement self-similar traffic generator using the aggregated On/Off source model, based on the analysis of the On-Off source model, FFT-FGN(Fast Fourier Transform-Fractional Gaussian Noise) model, and RMD(Random Midpoint Displacement) model. We confirmed the self-similarity of our generated traffic by checking the packet inter-arrival time of TCPdump data. Further we applied the implemented traffic generator to the performance evaluation of DiffServ networks and observed the effect of performance to the a value of the On/Off model, and performance of EF/BE class traffic by CBQ.

Microcomputer-Based Post-Processorfor Large Finite Element Analysis (대규모 유한요소해석에 활용되는 소형컴퓨터용 후처리 그래픽 프로그램)

  • 이성우;이선구;이태연
    • Computational Structural Engineering
    • /
    • v.2 no.4
    • /
    • pp.69-77
    • /
    • 1989
  • Until recently post-processing of finite element model has been heavily relied on expensive graphic peripheral devices. With the aid of inexpensive microcomputers, very economical post-processor graphics program called MICRO-POST has been developed. Model geometry or results of analysis for the unlimited meshes can be easily presented in a number of low-cost graphic devices. The paper presents the procedure obtaining the device-independent graphics, and the structure and functions of the program. It also describes efficient I/O scheme to overcome the memory limitation, and dialogue-type input technique to control the plot operation in an interactive manner. Through the post processing examples for the general purpose finite element programs, it demonstrates the usefulness of the program.

  • PDF

A Comparative Study on the REV, non-REV and Joint Network Methods for Analysis of Groundwater Flow in Jointed Rock Masses (절리암반내 지하수 유동해석을 위한 대표체적법, 비대표체적법 및 절리망 해석법의 비교 연구)

  • 문현구
    • Journal of the Korean Geotechnical Society
    • /
    • v.15 no.5
    • /
    • pp.217-228
    • /
    • 1999
  • The three methods of analysis (i) REV(representative elemental volume), (ii) non-REV and (iii) joint network analysis are introduced in this paper to analyze the groundwater flow in jointed rock mass and the inflow into underground excavations. The results from those methods are compared one another to reveal their characteristics by varying the number of joints and the diameter of the opening. The pre-processor, the so-called sequential analysis, is introduced to predict the equivalent hydraulic conductivity of a jointed rock mass having a number of intersecting joints. Using the finite element mesh, joint map and sequential analysis, the equivalent hydraulic conductivities are calculated for all 445 elements. The hydraulic inhomogeneity and the determination of the representative properties of jointed rock masses are discussed. In the REV analysis where the entire rock mass is homogenized through the representative properties, the inflow is increased regularly and consistently by increasing the joint density, the opening size and the conductivity contrast value. Though the non-REV analysis showed irregular variation of the inflow due to the local inhomogeneity allowed to individual elements, the inflow approached the REV results as the characteristic length increases. The joint network analysis showed the most sensitive reaction to the joint density, the opening size and the presence of the network crossing the opening. The reliability of the network analysis depends on the geometric data of individual joints. In view of the limited field data on joint geometry and possible uncertainty the REV and non-REV methods are considered more practical and rational than the joint network analysis.

  • PDF