Sharif Digital Repository

An integrated probabilistic framework is proposed with a set of probabilistic models that are compatible with the reliability methods used in resilience-based design (RBD). Previous seismic resilience estimation frameworks primarily have used conditional probabilities and total probability integration to compute the expected seismic resilience index while considering the effects of only a few types of uncertainty. However, a high level of uncertainty is associated with seismic resilience estimation. This study presents a probabilistic framework and models that can consider the effect of unlimited uncertainties when calculating the probability distribution of the seismic resilience index for... 

This paper puts forward a reliability-based approach to seismic design of steel moment frame structures isolated with lead-rubber bearing (LRB) devices. The system is modeled as an equivalent two-degree-of-freedom system in which the superstructure is assumed to remain elastic, while a bilinear behavior is assigned to the isolation level. Furthermore, the uncertainties associated with the superstructure properties including mass and stiffness are taken into account by adopting appropriate probability density functions. To design the base-isolated structure, the paper proposes “reliability curves” that given a target reliability, return the two key design parameters: the natural period of the... 

A library of probabilistic models for prediction of seismic risk is presented. The models are specifically intended for use with reliability methods to compute event probabilities, such as seismic loss probabilities. Several models are presented here for the first time. In particular, new and generic models are proposed for earthquake location, regional loss, building response, building damage, and building loss. Each model is presented with an explanation of its development and a discussion of its predictions. In addition, models from the literature are " smoothed" to make them amenable to reliability analysis. The models are implemented in a new computer program that is tailored for... 

This paper puts forward a framework for optimal mitigation of regional risk to enhance the resilience of civil infrastructure. To meet this objective, probabilistic models, methods, and software are developed and applied. The work is conducted within a new reliability-based approach, in which reliability methods compute risk. This contrasts several contemporary approaches for risk analysis. Risk, in this context, denotes the probability of exceeding monetary loss. Evaluating such probabilities requires probabilistic models for hazards, response, damage, and loss. This motivates the following contributions in this paper. First, a new computer program is developed that is tailored to conduct... 

This study presents a framework that utilizes Bayesian regression to create probabilistic cost models for retrofit actions. Performance improvement is the key parameter introduced in the proposed framework. The incorporation of this novel feature facilitates the characterization of retrofit cost as a continuous function of the desired performance improvement. Accounting for the performance gained from retrofit enables the use of the models in determining the optimal level of retrofit. Furthermore, accounting for the model uncertainty facilitates the use of the models in risk and reliability analyses. The proposed framework is applied to create seismic retrofit cost models for masonry school... 

Multi-label classification with a large set of labels is a challenging task. Label-Space Dimension Reduction (LSDR) is the most popular approach that addresses this problem. LSDR methods project the high-dimensional label vectors onto a low-dimensional space that can be predicted from the feature space. Many LSDR methods assume that the training data provide complete label vector for all training samples while this assumption is usually violated particularly when label vectors are high dimensional. In this paper, we propose a probabilistic model that has an effective mechanism to handle missing and noisy labels. In the proposed Bayesian network model, a set of auxiliary random variables,... 

A new time-domain hierarchical Bayesian framework is proposed to improve the performance of Bayesian methods in terms of reliability and robustness of estimates particularly for uncertainty quantification and propagation in structural dynamics. The proposed framework provides a reliable approach to account for the variability of the inference results observed when using different data sets. The proposed formulation is compared with a state-of-the-art Bayesian approach using numerical and experimental examples. The results indicate that the hierarchical Bayesian framework provides a more realistic account of the uncertainties, whereas the non-hierarchical Bayesian approach severely... 

A new time-domain hierarchical Bayesian framework is proposed to improve the performance of Bayesian methods in terms of reliability and robustness of estimates particularly for uncertainty quantification and propagation in structural dynamics. The proposed framework provides a reliable approach to account for the variability of the inference results observed when using different data sets. The proposed formulation is compared with a state-of-the-art Bayesian approach using numerical and experimental examples. The results indicate that the hierarchical Bayesian framework provides a more realistic account of the uncertainties, whereas the non-hierarchical Bayesian approach severely... 

Here, we investigate the influence of elevated temperatures with negligible ambient oxygen on mechanical properties of various embedded glass fiber reinforced polymer (GFRP) profiles, as well as the application of a predictive Bayesian model for predicting these properties. Both the flexural and compressive properties of FRP profiles were investigated through the tests of I-shaped and box-shaped profiles. To determine the impact of low and high elevated temperature, the profiles were exposed to a wide range of temperatures (i.e., 25–550 °C); effects of the exposure time were also investigated. Experiments showed that specimens exposed to higher elevated temperatures for longer time periods... 

Multi-label classification has received considerable interest in recent years. Multi-label classifiers usually need to address many issues including: handling large-scale datasets with many instances and a large set of labels, compensating missing label assignments in the training set, considering correlations between labels, as well as exploiting unlabeled data to improve prediction performance. To tackle datasets with a large set of labels, embedding-based methods represent the label assignments in a low-dimensional space. Many state-of-the-art embedding-based methods use a linear dimensionality reduction to map the label assignments to a low-dimensional space. However, by doing so, these... 

The total cost of earthquakes is in this paper modeled as a continuous random variable that includes the cost of damage and the cost of construction to prevent damage. Realizations of this variable are obtained by evaluating an array of probabilistic models that take many basic random variables as input. Consequently, analyses can be conducted to determine the mean cost, as well as cost exceedance probabilities and other measures of seismic risk. Such results are employed to address the underlying decision problem, namely to minimize the total cost of earthquakes when that cost is a continuous random variable. Applicable decision theories are outlined and several risk measures are... 

In this paper, the problem of multimodal soft coupling under the Bayesian framework when variance of probabilistic model is unknown is investigated. Similarity of shared factors resulted from Nonnegative Matrix Factorization (NMF) of multimodal data sets is controlled in a soft manner by using a probabilistic model. In previous works, it is supposed that the probabilistic model and its parameters are known. However, this assumption does not always hold. In this paper it is supposed that the probabilistic model is already known but its variance is unknown. So the proposed algorithm estimates the variance of the probabilistic model along with the other parameters during the factorization... 

Operational and planning studies of high-wind penetrated power systems have well come to the light as a major concern of future energy systems. This paper focuses on the procedure of determining required static reserve of the high-wind penetrated power systems which has not been well accompanied by comprehensive analysis and proper modeling tools. To reach this goal, first, a probabilistic algorithm has been proposed to effectively model the variations in output generation of wind turbines. In this algorithm, the fuzzy c-means clustering method (FCM) is exploited as an efficient as well as robust clustering method to find the multi-state model of wind turbines output generation. Based on... 

In this paper, a seismic reliability-based approach is proposed to design inelastic steel moment frame structures isolated by lead-rubber bearing (LRB) systems. An equivalent two-degree-of-freedom system is assumed in which a bilinear behaviour is assigned to both the superstructure and the base. Furthermore, uncertainties associated with the equivalent superstructure mass, stiffness, and yield properties are taken into account by employing proper probability density functions. The proposed design approach is twofold: 1) Reliability curves that return the key design parameters of the inelastic base-isolated structure including: the period of the superstructure, the target base displacement,... 

This paper presents probabilistic models and methods for detailed seismic risk analysis of structures using structural reliability methods. This approach to risk analysis is an alternative to those that employ the theorem of total probability and conditional probability distributions. Detailed risk analysis entails probabilistic quantification of responses, the ensuing damage of individual structural and nonstructural components, and the resulting economic and social losses. Such an analysis requires a library of probabilistic models for hazards, responses, damage, repair cost, downtime, and casualty with a specific format as presented in this paper. Two analysis options are proposed: one... 

Multi-label classification has received considerable interest in recent years. Multi-label classifiers usually need to address many issues including: handling large-scale datasets with many instances and a large set of labels, compensating missing label assignments in the training set, considering correlations between labels, as well as exploiting unlabeled data to improve prediction performance. To tackle datasets with a large set of labels, embedding-based methods represent the label assignments in a low dimensional space. Many state-of-the-art embedding-based methods use a linear dimensionality reduction to map the label assignments to a low-dimensional space. However, by doing so, these... 

The permeability of roller-compacted concrete (RCC) substantially affects its functionality and safety. This study investigates the effect of mix design parameters on the performance of RCC. For this purpose, approximately 500 laboratory specimens were prepared and tested. A formula and an artificial neural network (ANN) were proposed to predict the permeability coefficient of RCC by considering the main parameters, which were then verified independently using new specimens. Furthermore, the experimental data were analyzed by the Taguchi method and analysis of variance (ANOVA) to evaluate the level of parameter contribution. Based on the results, the permeability coefficient was highly...