Scholarlink

<p>This study presents a robust framework for automatic multi-objective performance-based seismic design (PBSD) of buildings, balancing computational efficiency and optimal design outcomes. PBSD aims to elevate structural resilience against seismic events by judiciously balancing safety and economic considerations. The overarching objective of PBSD is ideally addressed within a multi-objective optimization (MOO) framework, utilizing inelastic time-history analysis as the evaluation tool. Although the computational demands of such an approach are significant, prior research indicates that relying on approximate pushover analyses in the MOO process can lead to unreliable solutions. This paper introduces the extended fast converging multi-objective particle swarm optimization (FC-MOPSO) algorithm, which offers four alternative MOO procedures—namely, the Direct Nonlinear Dynamic Procedure (NDP), the Direct Linear Dynamic Procedure (LDP), the Surrogate LDP, and the Surrogate Nonlinear Static Procedure (NSP) models—to achieve reliable inelastic time-history-based optimal Pareto fronts within the PBSD framework. The effectiveness of the proposed models is demonstrated through the MOO design of three steel moment-resisting frames (MRFs) with setbacks. The results showcase substantial reductions in computational costs for the Direct and Surrogate LDP models compared to the Direct NDP model, thereby making them practical for engineering practitioners. Furthermore, the findings reveal that when the optimal design is selected through a Pareto front based on the two concurrent objectives—minimizing both the initial cost and the life-cycle cost—the Direct and Surrogate LDP models outperform the Surrogate NSP model.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Computationally Efficient Framework for Inelastic Multi-objective Performance-Based Seismic Design Using Surrogate Linear Dynamic Models

  • 期刊论文

  • 发表日期：

    2025年5月10日

查看全文
  • Vahid Mokarram,
  • Mohammad Reza Banan

摘要

This study presents a robust framework for automatic multi-objective performance-based seismic design (PBSD) of buildings, balancing computational efficiency and optimal design outcomes. PBSD aims to elevate structural resilience against seismic events by judiciously balancing safety and economic considerations. The overarching objective of PBSD is ideally addressed within a multi-objective optimization (MOO) framework, utilizing inelastic time-history analysis as the evaluation tool. Although the computational demands of such an approach are significant, prior research indicates that relying on approximate pushover analyses in the MOO process can lead to unreliable solutions. This paper introduces the extended fast converging multi-objective particle swarm optimization (FC-MOPSO) algorithm, which offers four alternative MOO procedures—namely, the Direct Nonlinear Dynamic Procedure (NDP), the Direct Linear Dynamic Procedure (LDP), the Surrogate LDP, and the Surrogate Nonlinear Static Procedure (NSP) models—to achieve reliable inelastic time-history-based optimal Pareto fronts within the PBSD framework. The effectiveness of the proposed models is demonstrated through the MOO design of three steel moment-resisting frames (MRFs) with setbacks. The results showcase substantial reductions in computational costs for the Direct and Surrogate LDP models compared to the Direct NDP model, thereby making them practical for engineering practitioners. Furthermore, the findings reveal that when the optimal design is selected through a Pareto front based on the two concurrent objectives—minimizing both the initial cost and the life-cycle cost—the Direct and Surrogate LDP models outperform the Surrogate NSP model.