Determining launching construction parameters of superlong superwide bridges: A multiobjective optimization method using machine learning techniques. (July 2022)

• Record Type: Journal Article
• Authors: Sun, Yuan; Tai, Xiao-xiao; Liu, Kai; Zhu, Ai-zhu; Zhu, Hong-ping
• Is Part Of: Structures. Volume 41(2022)
• Page Start: 15
• Page End: 28
• Publication Date: 2022-07
• Subjects: Construction -- Optimization -- Finite element -- Multiobjective -- Intelligent algorithm; Structural engineering -- Periodicals; 624.1
• DOI: 10.1016/j.istruc.2022.04.093
• Languages: English
• ISSNs: 2352-0124

Abstract

Highlights: A multiobjective optimization method is proposed for determining launching construction parameters of bridges. A surrogated-assisted optimization strategy using machine learning technique is employed. The GSA-SVR algorithm is used to predict the maximum local stresses of girders during construction. The method is successfully applied to the launching construction of a superlong superwide steel box girder bridge. Abstract: The launching construction parameters of bridges, such as the spacing of temporary piers and the length of launching noses, significantly affect the cost, shape and safety of the construction. Determining such parameters generally requires multistage manual comparison studies upon finite element (FE) analysis, which might become extremely tedious when superlong superwide structures with local stress concerns are involved. This paper proposes a multiobjective optimization method for determining the launching construction parameters of bridges while considering local stress constraints. The multiobjective particle swarm optimization (MOPSO) is employed as an optimization tool, whereas the intelligent algorithm (GSA-SVR) is used to establish a surrogate model. This model creates a direct mapping relationship between the design variables and the most adverse local stresses of deck segments in construction. Then, the launching construction of a superlong superwide steel box girder bridge is introduced as a case study. The effectiveness and merits of the method are validated by the optimization results and parametric studies. The study shows how the construction parameters can affect the trade-off between the cost and the shape results..