Algoritmo genético modificado aplicado à otimização de vigas protendidas
Prestressing has become increasingly attractive, as it allows for larger spans, greater architectural flexibility, faster execution, and greater durability. All these advantages can make prestressed concrete structures more economical when combined with an efficient scaffolding system and formwork s...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Estado: | Versión publicada |
| Fecha de publicación: | 2025 |
| País: | Brasil |
| Institución: | Universidade Federal do Ceará (UFC) |
| Repositorio: | Repositório Institucional da Universidade Federal do Ceará (UFC) |
| Idioma: | portugués |
| OAI Identifier: | oai:repositorio.ufc.br:riufc/83067 |
| Acceso en línea: | http://repositorio.ufc.br/handle/riufc/83067 |
| Access Level: | acceso abierto |
| Palabra clave: | CNPQ::ENGENHARIAS::ENGENHARIA CIVIL Algoritmos genéticos Vigas de concreto protendido Operador de cruzamento Otimização estrutural Genetic algorithms Prestressed concrete beams Crossover operator Structural optimization |
| Sumario: | Prestressing has become increasingly attractive, as it allows for larger spans, greater architectural flexibility, faster execution, and greater durability. All these advantages can make prestressed concrete structures more economical when combined with an efficient scaffolding system and formwork system. The designer can appropriately define the dimensions of beams and slabs, as well as parameters such as the quantity and layout of tendons based on their experience. However, optimization techniques can be used to find an optimal solution, where design parameters can be determined so that the solution minimizes, for example, the cost, composed of the cost of materials—which includes the volume of concrete, weight of cables, and passive reinforcement—and the cost-of-service execution. Genetic Algorithms (GAs) have been widely used to optimize various types of problems, possessing control parameters and operators that are generally problem-dependent, requiring calibrations calibrations in each case. Thus, some GA variants have been proposed in the literature, but with few applications to prestressed concrete structure design problems. In this work, existing and newly proposed GA formulations for the crossover operator are evaluated for their efficiency and effectiveness in the design of prestressed beams. Structural analyses were performed using the FAST (Finite Element Analysis Tool) software, and the new formulations were implemented in the genetic algorithm of the BIOS (Biologically Inspired Optimization System) software. Both open-source programs were developed in C++ by Laboratório de Mecânica Computacional e Visualização (LMCV) at Universidade Federal do Ceará Ceará (UFC). The proposed algorithms were compared with crossover alternatives in the literature and tested using mathematical benchmarks. Afterward, they were applied to multi-span prestressed beam problems. The effect of varying the fck on the final cost of the optimized beam was analyzed. In summary, the results showed that the proposed operators outperform the alternatives in the literature. |
|---|