Optimum Design of Geometrically Nonlinear Suspended and Curved Pedestrian Bridges Using Swarm Intelligence Techniques

Abstract

Abstract : This study aims to achieve the optimum design of a curved footbridge hanging by a curved arch. For this purpose, the design of the I-4 Maitland pedestrian bridge located in the Florida was considered, and the structural of the footbridge was modeled using SAP2000. Once it is designed, the structure members are classified into seven groups per their functionality and modality for carrying the design loads; however, these groups are designated as standard Wsection, hollow structural sections (HSS), cables and pipe sections. The optimization uses the cross-sectional area of the classified sections as variables due to their effect on the overall weight of the footbridge structure and their criteria are evaluated by AASHTO / AISC standards. Three metaheuristic algorithms are used in the optimization process which are: learning-based optimization teaching algorithm (TLBO), artificial bee colony algorithm (ABC), and differential evolution algorithm (DE). The MATLAB software is used for maintaining optimization process, along with API SM toolbox considering it as a translator between SAP2000 and MATLAB. The results show that the DE algorithm is best optimizer for the design as the design weight was reduces up to 27.79% of its initial weight in just 11.59 days, while the ABC algorithm reaches 33.86% in 21 days and the TLBO algorithm reaches 27.79%. taking 30.5 days.