基于响应面法的刚构桥合龙顶推模型修正

Correction of Rigid Frame Bridge Closure and Incremental Launching Model Based on Response Surface Method

摘要:

为解决多跨连续刚构桥实桥结构的力学性能与依照设计图纸建立的有限元模型存在较大差异的问题，提出一种结合响应面法和实桥试验的模型修正方法：首先根据敏感性分析选取待修正参数，再采用中心复合设计方法来获取试验样本；接着运用F检验法来选取影响显著的自变量参数作为响应面函数的参数项，建立二次多项式响应面函数，以减小墩顶顶推位移计算值与实测值的差异作为目标进行优化。结果表明：得到修正后模型的墩顶位移计算值与实测值吻合良好（与实测值的相对误差由最高35.7%降至全部小于3%），表明修正后的有限元模型能高精度地模拟桥梁实际结构在合龙顶推阶段的力学性能，为后续求解多跨连续刚构桥的合理顶推力组合提供有效的模型基础。

Abstract:

To address the significant discrepancy between the mechanical performance of the actual multi-span continuous rigid frame bridge structure and the finite element model established based on the design drawings, a model correction method that combines the response surface method with the actual bridge test is proposed. Firstly, the parameters to be corrected are selected based on the sensitivity analysis, and then the central composite design method is used to obtain the test samples. Subsequently, the F-test method is employed to select the significant independent variable parameters as the parameter items of the response surface function. A quadratic polynomial response surface function is established to optimize it with the goal of reducing the difference between the calculated value and the measured value of the pier top incremental launching displacement. The results show that the calculated values of the pier top displacement of the corrected model are in good agreement with the measured values (the relative error with the measured values is decreased from a maximum of 35.7% to all less than 3%), indicating that the corrected finite element model can accurately simulate the mechanical performance of the actual bridge structure during the closure and incremental launching stages, which provides an effective model basis for the subsequent solution of the reasonable incremental launching force combination of the multi-span continuous rigid frame bridge.