Weight Optimization of Steel Plate Girders: A Particle Swarm Optimization Framework with ANSI/AISC 360-16 Compliance

Abstract

This paper presents a framework for designing steel plate girders according to ANSI/AISC 360-16 specifications using Particle Swarm Optimization (PSO) with adaptive penalty. The objective is to minimize the total weight of the girder system while satisfying flexural strength, shear strength, and serviceability requirements. The optimization considers key design variables beam depth, flange width, web thickness, flange thickness, and transverse stiffener arrangements. Additionally, manufacturing feasibility is considered by optimizing dimensions and using standard steel plate thicknesses. Comparative testing results demonstrate that the developed PSO method can reduce the weight of steel plate girders by 2.7% and 9.0% compared to conventional design methods and other optimization algorithms, respectively, under the same constraints.