• Atmosphere
• /
• v.30 no.4
• /
• pp.405-420
• /
• 2020

In this study, we investigate the predictability of intraseasonal monthly temperature and precipitation variations using hindcast datasets from eight global circulation models participating in the operational multi-model ensemble (MME) seasonal prediction system of the Asia-Pacific Economic Cooperation Climate Center for the 1983~2010 period. These intraseasonal monthly variations are defined by categorical deterministic analysis. The monthly temperature and precipitation are categorized into above normal (AN), near normal (NN), and below normal (BN) based on the σ-value ± 0.43 after standardization. The nine patterns of intraseasonal monthly variation are defined by considering the changing pattern of the monthly categories for the three consecutive months. A deterministic and a probabilistic analysis are used to define intraseasonal monthly variation for the multi-model consisting of numerous ensemble members. The results show that a pattern (pattern 7), which has the same monthly categories in three consecutive months, is the most frequently occurring pattern in observation regardless of the seasons and variables. Meanwhile, the patterns (e.g., patterns 8 and 9) that have consistently increasing or decreasing trends in three consecutive months, such as BN-NN-AN or AN-NN-BN, occur rarely in observation. The MME and eight individual models generally capture pattern 7 well but rarely capture patterns 8 and 9.

• Smart Structures and Systems
• /
• v.32 no.1
• /
• pp.61-81
• /
• 2023

This study presents an ensemble learning based Bayesian model updating approach for structural damage diagnosis. In the developed framework, the structure is initially decomposed into a set of substructures. The autoregressive moving average (ARMAX) model is established first for structural damage localization based structural motion equation. The wavelet packet decomposition is utilized to extract the damage-sensitive node energy in different frequency bands for constructing structural surrogate models. Four methods, including Kriging predictor (KRG), radial basis function neural network (RBFNN), support vector regression (SVR), and multivariate adaptive regression splines (MARS), are selected as candidate structural surrogate models. These models are then resampled by bootstrapping and combined to obtain an ensemble model by probabilistic ensemble. Meanwhile, the maximum entropy principal is adopted to search for new design points for sample space updating, yielding a more robust ensemble model. Through the iterations, a framework of surrogate ensemble learning based model updating with high model construction efficiency and accuracy is proposed. The specificities of the method are discussed and investigated in a case study.

• Coupled systems mechanics
• /
• v.12 no.5
• /
• pp.429-444
• /
• 2023

The comprehension and structural modeling of masonry constructions is fundamental to safeguard the integrity of built cultural assets and intervene through adequate actions, especially in earthquake-prone regions. Despite the availability of several modeling strategies and modern computing power, modeling masonry remains a great challenge because of still demanding computational efforts, constraints in performing destructive or semi-destructive in-situ tests, and material uncertainties. This paper investigates the shear behavior of masonry walls by applying a plane-stress FE continuum model with the Modified Masonry-like Material (MMLM). Epistemic uncertainty affecting input parameters of the MMLM is considered in a probabilistic framework. After appointing a suitable probability density function to input quantities according to prior engineering knowledge, uncertainties are propagated to outputs relying on gPCE-based surrogate models to considerably speed up the forward problem-solving. The sensitivity of the response to input parameters is evaluated through the computation of Sobol' indices pointing out the parameters more worthy to be further investigated, when dealing with the seismic assessment of masonry buildings. Finally, masonry mechanical properties are calibrated in a probabilistic setting with the Bayesian approach to the inverse problem based on the available measurements obtained from the experimental load-displacement curves provided by shear compression in-situ tests.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.18 no.3
• /
• pp.151-160
• /
• 2014

Probabilistic tsunami hazard analysis (PTHA) is based on the approach of probabilistic seismic hazard analysis (PSHA) which is performed using various seismotectonic models and ground-motion prediction equations. The major difference between PTHA and PSHA is that PTHA requires the wave parameters of tsunami. The wave parameters can be estimated from tsunami propagation analysis. Therefore, a tsunami simulation analysis was conducted for the purpose of evaluating the wave parameters required for the PTHA of Uljin nuclear power plant (NPP) site. The tsunamigenic fault sources in the western part of Japan were chosen for the analysis. The wave heights for 80 rupture scenarios were numerically simulated. The synthetic tsunami waveforms were obtained around the Uljin NPP site. The results show that the wave heights are closely related with the location of the fault sources and the associated potential earthquake magnitudes. These wave parameters can be used as input data for the future PTHA study of the Uljin NPP site.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.6 no.5
• /
• pp.649-656
• /
• 2011

This study is a numerical representative modelling analysis for applying the process that unravels networks between cells in genetics to Network of informatics. Using the probabilistic graphical model, the insight from the data describing biological networks is used for making a probabilistic function. Rather than a complex network of cells, we reconstruct a simple lower-stage model and show a genetic representation level from the genetic based network logic. We made probabilistic graphical models from genetic data and extend them to genetic representation data in the method of network modelling in informatics.

• Computers and Concrete
• /
• v.3 no.2_3
• /
• pp.123-144
• /
• 2006

Reinforced concrete corbels are structural elements widely used in practical engineering. The complex response of these elements is described in design codes in a simplified manner. These formulations are not sufficient to show the real behavior, which, however, is an essential prerequisite for the manufacturing of numerous elements. Therefore, a deterministic and probabilistic study has been performed, which is described in this contribution. Real complex structures have been modeled by means of the finite element method supported primarily by experimental works. The main objective of this study was the detection of uncertainties effects and safety margins not captured by traditional codes. This aim could be fulfilled by statistical considerations applied to the investigated structures. The probabilistic study is based on advanced Monte Carlo simulation techniques and sophisticated nonlinear finite element formulations.

• Steel and Composite Structures
• /
• v.33 no.5
• /
• pp.641-652
• /
• 2019

This paper investigates the effect of aftershocks on the seismic performance of self-centering (SC) prestressed concrete frames using the probabilistic seismic demand analysis methodology. For this purpose, a 4-story SC concrete frame and a conventional reinforced concrete (RC) frame are designed and numerically analyzed through nonlinear dynamic analyses based on a set of as-recorded mainshock-aftershock seismic sequences. The peak and residual story drifts are selected as the demand parameters. The probabilistic seismic demand models of the SC and RC frames are compared, and the SC frame is found to have less dispersion of peak and residual story drifts. The results of drift demand hazard analyses reveal that the SC frame experiences lower peak story drift hazards and significantly reduced residual story drift hazards than the RC frame when subjected to the mainshocks only or the mainshock-aftershock sequences, which demonstrates the advantages of the SC frame over the RC frame. For both the SC and RC frames, the influence of as-recorded aftershocks on the drift demand hazards is small. It is shown that artificial aftershocks can produce notably increased drift demand hazards of the RC frame, while the incremental effect of artificial aftershocks on the drift demand hazards of the SC frame is much smaller. It is also found that aftershock polarity does not influence the drift demand hazards of both the SC and RC frames.

• Wind and Structures
• /
• v.30 no.4
• /
• pp.405-421
• /
• 2020

Quayside container cranes are important delivery machineries located in the most frontiers of container terminals, where strong wind attacks happen occasionally. Since the previous researches on quayside container cranes mainly focused on the mean wind load and static response characteristics, the fluctuating wind load and dynamic response characteristics require further investigations. In the present study, the aerodynamic wind loads on quayside container cranes were obtained from wind tunnel tests. The probabilistic and spectral models of the fluctuating aerodynamic loads were established. Then the joint probabilistic distributions of dynamic wind-induced responses were derived theoretically based on a series of Gaussian and independent assumption of resonant components. Finally, the results were validated by time domain analysis using wind tunnel data. It is concluded that the assumptions are acceptable. And the presented approach can estimate peak dynamic sliding force, overturning moments and leg uplifts of quayside container cranes effectively and efficiently.

• Nuclear Engineering and Technology
• /
• v.44 no.5
• /
• pp.471-482
• /
• 2012

To assess the risk of nuclear power plant operation and to determine the risk impact of digital systems, there is a need to quantitatively assess the reliability of the digital systems in a justifiable manner. The Probabilistic Risk Analysis (PRA) is a tool which can reveal shortcomings of the NPP design in general and PRA analysts have not had sufficient guiding principles in modelling particular digital components malfunctions. Currently digital I&C systems are mostly analyzed simply and conventionally in PRA, based on failure mode and effects analysis and fault tree modelling. More dynamic approaches are still in the trial stage and can be difficult to apply in full scale PRA-models. As basic events CPU failures, application software failures and common cause failures (CCF) between identical components are modelled.The primary goal is to model dependencies. However, it is not clear which failure modes or system parts CCF:s should be postulated for. A clear distinction can be made between the treatment of protection and control systems. There is a general consensus that protection systems shall be included in PRA, while control systems can be treated in a limited manner. OECD/NEA CSNI Working Group on Risk Assessment (WGRisk) has set up a task group, called DIGREL, to develop taxonomy of failure modes of digital components for the purposes of PRA. The taxonomy is aimed to be the basis of future modelling and quantification efforts. It will also help to define a structure for data collection and to review PRA studies.

• Nuclear Engineering and Technology
• /
• v.52 no.1
• /
• pp.211-221
• /
• 2020

One fundamental element of probabilistic safety assessment (PSA) is the initiating event (IE) analysis. Since IE frequencies can change over time, time-trend analysis is required to obtain optimized IE frequencies. Accordingly, such time-trend analyses have been employed to estimate industry-average IE frequencies for use in the PSAs of U.S. nuclear power plants (NPPs); existing PSAs of Korean NPPs, however, neglect such analysis in the estimation of IE frequencies. This article therefore provides the method for and results of estimating Korean industry-average IE frequencies using time-trend analysis. It also examines the effects of the IE frequencies obtained from this study on risk insights by applying them to recently updated internal events Level 1 PSA models (at-power and shutdown) for an OPR-1000 plant. As a result, at-power core damage frequency decreased while shutdown core damage frequency increased, with the related contributions from each IE category changing accordingly. These results imply that the incorporation of time-trend analysis leads to different IE frequencies and resulting risk insights. The IE frequency distributions presented in this study can be used in future PSA updates for Korean NPPs, and should be further updated themselves by adding more recent data.