• Computers and Concrete
• /
• 제13권5호
• /
• pp.659-677
• /
• 2014

A Bayesian response surface updating procedure is applied in order to update the parameters of the covariance function of a random field for concrete properties based on a limited number of available measurements. Formulas as well as a numerical algorithm are presented in order to update the parameters of response surfaces using Markov Chain Monte Carlo simulations. The parameters of the covariance function are often based on some kind of expert judgment due the lack of sufficient measurement data. However, a Bayesian updating technique enables to estimate the parameters of the covariance function more rigorously and with less ambiguity. Prior information can be incorporated in the form of vague or informative priors. The proposed estimation procedure is evaluated through numerical simulations and compared to the commonly used least square method.

• 산업공학
• /
• 제24권2호
• /
• pp.119-127
• /
• 2011

Most companies have been increasing temporary work projects to maximize the usage of their resources. They also have been developing the effective techniques for analyzing and managing the state of the projects. In order to monitor the state of a project in real-time and predict the project's future state more accurately, this paper suggests the Bayesian Network (BN) as a tool for discovering the causes of project risk and presenting the failure probability of the project. The proposed BN modeling method with consideration of the Earned Value Management (EVM) method shows how to induce the predictive and conditional probability of the risk occurrence in the future. The advantages of the suggested model are (1) that the cause of a project risk can be easily figured out via the BN, (2) that the future value of the project can be sufficiently increased by updating relevant components of the project, and (3) that more credible prediction can be made in the similar and future situation by using the data obtained in current analysis. A numerical example is also given.

• 한국지능시스템학회논문지
• /
• 제26권5호
• /
• pp.335-342
• /
• 2016

기 구축되어있는 베이지안 네트워크에서 다이나믹한 환경 변화가 발생 할 때, 관련된 베이지안 네트워크의 파라미터는 새롭게 형성된 데이터의 패턴에 적응하여 새로운 파라미터로 변경되어야 한다. 이때, 새로운 파라미터는 베이지안 네트워크의 인과관계를 고려하여 변경되어야 한다. 본 논문에서는 Expectation Maximization(EM)알고리즘과 Meta-Heuristics 기법 중 하나인 Harmony Search(HS)알고리즘을 이용한 다이나믹한 파라미터 업데이트 프레임웍을 제안한다. 일반적으로, EM 알고리즘은 숨겨진 파라미터를 추정하는데 유효한 알고리즘이지만 지역 최적값에 수렴한다는 단점을 가지고 있다. 이 문제를 해결하기 위해서 본 논문은 Maximum Likelihood Estimator(MLE)의 파라미터가 글로벌 최적값을 지향하도록 하기위하여 메타휴리스틱 방법론의 하나인 HS를 적용한다. 제안된 방법은 EM 알고리즘의 단점을 보완하고 글로벌 최적값에 수렴하는 MLE의 파라미터를 추정하여 다이나믹하게 변화하는 환경에서도 사용 가능한 베이지안 네트워크의 학습 및 전파프레임웍을 제시한다.

• Structural Monitoring and Maintenance
• /
• 제2권1호
• /
• pp.77-93
• /
• 2015

The idea of using measured dynamic characteristics for damage detection is attractive because it allows for a global evaluation of the structural health and condition. However, vibration-based damage detection for complex structures such as long-span cable-supported bridges still remains a challenge. As a suspension or cable-stayed bridge involves in general thousands of structural components, the conventional damage detection methods based on model updating and/or parameter identification might result in ill-conditioning and non-uniqueness in the solution of inverse problems. Alternatively, methods that utilize, to the utmost extent, information from forward problems and avoid direct solution to inverse problems would be more suitable for vibration-based damage detection of long-span cable-supported bridges. The auto-associative neural network (ANN) technique and the probabilistic neural network (PNN) technique, that both eschew inverse problems, have been proposed for identifying and locating damage in suspension and cable-stayed bridges. Without the help of a structural model, ANNs with appropriate configuration can be trained using only the measured modal frequencies from healthy structure under varying environmental conditions, and a new set of modal frequency data acquired from an unknown state of the structure is then fed into the trained ANNs for damage presence identification. With the help of a structural model, PNNs can be configured using the relative changes of modal frequencies before and after damage by assuming damage at different locations, and then the measured modal frequencies from the structure can be presented to locate the damage. However, such formulated ANNs and PNNs may still be incompetent to identify damage occurring at the deck members of a cable-supported bridge because of very low modal sensitivity to the damage. The present study endeavors to enhance the damage identification capability of ANNs and PNNs when being applied for identification of damage incurred at deck members. Effort is first made to construct combined modal parameters which are synthesized from measured modal frequencies and modal shape components to train ANNs for damage alarming. With the purpose of improving identification accuracy, effort is then made to configure PNNs for damage localization by adapting the smoothing parameter in the Bayesian classifier to different values for different pattern classes. The performance of the ANNs with their input being modal frequencies and the combined modal parameters respectively and the PNNs with constant and adaptive smoothing parameters respectively is evaluated through simulation studies of identifying damage inflicted on different deck members of the double-deck suspension Tsing Ma Bridge.