Automated Approach For Multi-objective Optimization Of Steel Trusses Using Genetic Algorithms and Reliability

Resumo

Optimization is a valuable tool in structural engineering, enabling the achieve of optimal solutions that meet design constraints and objectives. The application of multi-objective optimization in steel trusses can bring significant benefits in terms of structural efficiency and economy. In this article, an automated approach was developed to obtain parametric and multi-objective optimal settings for flat truss structures. The methodology was developed based on the utilization of the finite element method for structural analysis and the integration of evolutionary computing techniques, employing Genetic Algorithms (GAs) via MATLAB. Furthermore, to address the uncertainties inherent in the optimization process, a reliability analysis was conducted using the First-Order Reliability Method (FORM). This analysis took into consideration the variability of key parameters such as area, density, diameters, thicknesses and displacement, treating them as random variables. The results demonstrated good accuracy when compared to benchmark trusses from the literature, highlighting the potential of applying multi-objective optimization in steel trusses to improve efficiency and quality of these structures.