TY - BOOK

T1 - Simulation methods for reliability-based design optimization and model updating of civil engineering structures and systems

AU - Jerez Urquieta, Danko José

PY - 2023

Y1 - 2023

N2 - This thesis presents a collection of original contributions pertaining to the subjects of reliability-based design optimization (RBDO) and model updating of civil engineering structures and systems. In this regard, probability theory concepts and tools are instrumental in the formulation of the herein reported developments. Firstly, two approaches are devised for the RBDO of structural dynamical systems under stochastic excitation. Namely, a stochastic search technique is proposed for constrained and unconstrained RBDO problems involving continuous, discrete and mixed discrete-continuous design spaces, whereas an efficient sensitivity assessment framework for linear stochastic structures is implemented to identify optimal designs and evaluate their sensitivities. Moreover, two classes of model updating problems are considered. In this context, the Bayesian interpretation of probability theory plays a key role in the proposed solution schemes. Specifically, contaminant source detection in water distribution networks is addressed by resorting to a sampling-based Bayesian model class selection framework. Furthermore, an effective strategy for Bayesian model updating with structural reliability methods is presented to treat identification problems involving structural dynamical systems, measured response data, and high-dimensional parameter spaces. The approaches proposed in this thesis integrate stochastic simulation techniques as an essential part of their formulation, which allows obtaining non-trivial information about the systems of interest as a byproduct of the solution processes. Overall, the findings presented in this thesis suggest that the reported methods can be potentially adopted as supportive tools for a number of practical decision-making processes in civil engineering.

AB - This thesis presents a collection of original contributions pertaining to the subjects of reliability-based design optimization (RBDO) and model updating of civil engineering structures and systems. In this regard, probability theory concepts and tools are instrumental in the formulation of the herein reported developments. Firstly, two approaches are devised for the RBDO of structural dynamical systems under stochastic excitation. Namely, a stochastic search technique is proposed for constrained and unconstrained RBDO problems involving continuous, discrete and mixed discrete-continuous design spaces, whereas an efficient sensitivity assessment framework for linear stochastic structures is implemented to identify optimal designs and evaluate their sensitivities. Moreover, two classes of model updating problems are considered. In this context, the Bayesian interpretation of probability theory plays a key role in the proposed solution schemes. Specifically, contaminant source detection in water distribution networks is addressed by resorting to a sampling-based Bayesian model class selection framework. Furthermore, an effective strategy for Bayesian model updating with structural reliability methods is presented to treat identification problems involving structural dynamical systems, measured response data, and high-dimensional parameter spaces. The approaches proposed in this thesis integrate stochastic simulation techniques as an essential part of their formulation, which allows obtaining non-trivial information about the systems of interest as a byproduct of the solution processes. Overall, the findings presented in this thesis suggest that the reported methods can be potentially adopted as supportive tools for a number of practical decision-making processes in civil engineering.

U2 - 10.15488/14122

DO - 10.15488/14122

M3 - Doctoral thesis

CY - Hannover

ER -