Comparative Analysis of Multi-Objective and Single-Objective Optimization Approaches in Structural Engineering
Year: 2024
Type: conference-paper
Abstract: This study investigates the effectiveness of multi-objective optimization versus singleobjective optimization in structural engineering design.Through a comparative analysis, employing the same objective functions in both approaches across various scenarios, we assess their performance in balancing conflicting objectives while maintaining solution constraints.Single-objective optimization strategies involve formulating constraints based on one objective or constructing a scalar objective function with a single weight coefficient.Our findings reveal that while both approaches yield similar results, they differ significantly in complexity.Multi-objective optimization poses challenges in balancing competing objectives, while single-objective optimization with scalarization requires careful construction of the scalar objective function and weight parameter selection.However, single-objective optimization simplifies the optimization process when one objective is reduced to constraints.Additionally, the inclusion of auxiliary objective functions aids in solution refinement.Overall, our analysis highlights the potential for employing single-objective optimization as an alternative to multi-objective optimization, facilitating problem definition and enabling the incorporation of auxiliary objectives for enhanced optimization outcomes.
Source: Civil-comp conferences
DOI: 10.4203/ccc.9.3.3
Authors: B. Miller, L. Ziemiański
Institution: Rzeszów University of Technology
Language: English
Subfield: Computational Theory and Mathematics
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4401882091