Optimization of uncertain structures subject to stochastic wind loads under system-level first excursion constraints: A data-driven approach

• Published in: Computers & structures Vol. 210; pp. 58 - 68
• Main Authors: Suksuwan, Arthriya, Spence, Seymour M.J.
• Format: Journal Article
• Language: English
• Published: New York Elsevier Ltd 01.11.2018; Elsevier BV
• Subjects: Algorithms; First excursion constraints; High-dimensional problems; Monte Carlo simulation; Optimization; Reliability-based design optimization; Stochastic wind loads; System effectiveness; Wind; Wind engineering; Wind loads; Stochastic wind loads; Wind engineering; First excursion constraints; Reliability-based design optimization; High-dimensional problems; Monte Carlo simulation
• ISSN: 0045-7949; 1879-2243
• DOI: 10.1016/j.compstruc.2018.09.001

•A data-driven system-level reliability optimization framework is proposed.•The framework is specifically developed for large-scale wind-excited systems.•The dynamic wind loads are characterized through data-driven stochastic models.•A new decoupling method is proposed for solving the stochastic optimization problem.•Numerical examples are presented demonstrating the applicability of the method. This work proposes a novel data-driven optimization strategy that can efficiently handle system-level first excursion performance constraints posed on large-scale uncertain structures subject to general stochastic wind excitation. The framework is centered on defining and solving a limited sequence of decoupled optimization sub-problems. In particular, each problem is formulated in terms of information obtained from a single simulation carried out in the solution of the previous sub-problem. Two examples involving the optimal design of uncertain systems subject to stochastic wind loads are presented to demonstrate the effectiveness, efficiency, and scalability of the proposed framework.