桥梁结构静力模型修正响应面方法试验设计研究

Experimental Designs Study of Bridge Structure Statics-Load Test Model Updating Based Response Surface Methodology

在基于响应面方法的结构模型修正中,参数试验设计对于响应面模型(函数,)以及最后结构模型修正结果的精度具有决定作用.对正交、中心复合、均匀试验设计方法进行了比较和分析,并结合一片两跨连续梁结构静力试验模型修正的数值算例,分析了不同试验设计对结构模型修正结果的影响.理论和数值算例分析表明,均匀试验设计具有设计点(参数选取)较少、结构有限元计算量少,且精度可接受的优点,更适合基于响应面方法的桥梁结构模型修正.

Parameter design of experiments is essential importance for response surface model/function and result precision of structure model updating, in the process of bridge structure model updating based response surface methodology （RSM）. Orthogonal, Central composite and Uniform experimental designs are analyzed and compared. And with application on a two spans continuous beam numerical examples analysis, effect of structure model updating resuh by different experimental designs is calculated. Theory and the numerical examples analysis indicate that the resuh precision by Uniform experimental designs is acceptable, the parameter design pots and the structure finite-element calculation times can be reduced, and it more suit for bridge structure model updating based RSM.