Búsqueda exhaustiva por entornos aplicada al diseño económico de bóvedas de hormigón armado
[EN] This paper presents two gradient algorithms applied to the economic optimization of reinforced concrete vaults, typically used in the construction of underpasses. The algorithms are gradient multi start neighbourhood exhaustive search procedures. The vault is defined by 45 design variables and...
| Authors: | , , |
|---|---|
| Format: | article |
| Publication Date: | 2011 |
| Country: | España |
| Institution: | Universitat Politècnica de València (UPV) |
| Repository: | RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia |
| Language: | Spanish |
| OAI Identifier: | oai:riunet.upv.es:10251/46925 |
| Online Access: | https://riunet.upv.es/handle/10251/46925 |
| Access Level: | Open access |
| Keyword: | Road vaults Structural design Heuristic optimization Gray coding Concrete structures Bóvedas Diseño estructural Optimización heurística Código Gray Hormigón estructural INGENIERIA DE LA CONSTRUCCION |
| Summary: | [EN] This paper presents two gradient algorithms applied to the economic optimization of reinforced concrete vaults, typically used in the construction of underpasses. The algorithms are gradient multi start neighbourhood exhaustive search procedures. The vault is defined by 45 design variables and the objective function is an economic one. Both methods have been applied to a vault of 12.40 m of diameter and 3.00 m of lateral walls. Design variables have been coded in base 2 for algorithms GB1 and in Gray coding for GB2. The neighbourhood is defined by the set of solutions that differ in one bit. The Gray coding solves the lack of proximity between two solutions typical of the ordinary binary coding. The positive effect of the Gray coding is proven in the present paper, where the average cost of 3,000 runs of GB2 improves by 3.81% a similar run by algorithm GB1. In addition, GB2 attains the best cost solution. The paper includes a stop criterion for the algorithm based on the stability of the statistics of the multi start results. The algorithms are simple and can be applied to other structural problems. The structure of best cost has a high slenderness and a span to depth ratio of 40. The study reports savings of 6% when compared to a design by an experienced practitioner office. © 2010 CIMNE (UPC). Published by Elsevier España, S.L. All rights reserved. |
|---|