摘要： •A robust geotechnical design (RGD) methodology for braced excavations is presented.•RGD aims to identify optimal designs that are safe, robust, and cost-efficient.•Multi-objective optimization is performed to establish Pareto Front.•Pareto Front defines a trade-off relationship between cost and robustness.•Knee point concept is used for selecting most-preferred design from Pareto Front. In this paper, the authors present a methodology for the robust geotechnical design (RGD) of braced excavations in clayey soils. The maximum wall deflection induced by the excavation was chosen as the response of concern in the design and was computed using a finite element analysis model based upon the beam-on-elastic-foundation theory. The variation of the maximum wall deflection of a given design of a braced excavation due to uncertainty in the soil parameters and the surcharges was used as a measure of the design robustness. The robust design of the braced excavation system (including soil, wall, and support) was then formulated as a multi-objective optimization problem, in which the variation of the maximum wall deflection (a signal of the design robustness) and the cost were optimized with the strict safety constraints. Using a multi-objective genetic algorithm, the optimal designs were then determined, the results of which were presented as a Pareto Front that exhibited a trade-off relationship useful for design decision-making. Furthermore, the “knee point” concept, based upon the “gain-sacrifice” trade-off is used in the selection of the most-preferred design from the Pareto Front. Finally, a design example of a braced excavation system was used to illustrate the significance of this proposed methodology. 出版者： Amsterdam: Elsevier Ltd 出版日期： 2014-01-01 出處： Structural safety, 2014-01, Vol.49, p.37-44 版權： 2013 Elsevier Ltd 版權： 2015 INIST-CNRS 識別號： ISSN: 0167-4730 識別號： EISSN: 1879-3355 識別號： DOI: 10.1016/j.strusafe.2013.05.003 識別號： CODEN: STSADI
