Multiobjective Optimization for Bridge and Viaduct Design

Resumo

This study addresses the multiobjective optimization of bridge and viaduct design, incorporating three objectives: minimization of construction costs, reduction of environmental impacts through decreased CO₂ emissions, and maximization of structural durability. Viaducts play a critical role in Brazil's transportation infrastructure, making efficient and sustainable design solutions essential. Traditional bridge design predominantly relies on empirical and experience-based approaches, often resulting in suboptimal solutions regarding cost, sustainability, and structural performance. To overcome these limitations, this research proposes an optimization methodology based on heuristic metaheuristic algorithms: Multiple Objective Particle Swarm Optimization (MOPSO), Nondominated Sorting Genetic Algorithm II (NSGA-II), and Strength Pareto Evolutionary Algorithm 2 (SPEA2). The comparative analysis demonstrated the superior performance of MOPSO, as measured by hypervolume and solution diversity metrics. Subsequently, MOPSO parameters were calibrated using the Taguchi method to enhance solution quality, algorithm robustness, and computational efficiency. Parameter calibration significantly improved Pareto-optimal solutions, highlighting the importance of algorithmic fine-tuning. The effectiveness of the proposed methodology was validated through two real-world case studies of viaducts along the BR-376 highway in Atalaia and Mandaguaçu, Paraná, Brazil. Results showed substantial improvements compared to original structural designs, achieving construction cost reductions between 9.1% and 23.2%, CO₂ emissions reductions from 8.9% to 23.5%, and significant increases in structural lifespan, reaching up to 540.7%. Additionally, analysis of relationships between design variables and objectives revealed that larger cross-sectional dimensions typically resulted in lower construction costs, while smaller dimensions correlated with reduced environmental impacts. The developed optimization approach offers engineers a practical, systematic tool for balanced structural solutions that meet contemporary sustainability, economic feasibility, and durability demands. The software developed, including its source code, is freely accessible at https://www.optimusviaduto.ufscar.br/.