利用Kriging方法进行结构模型修改

结构模型修改已经演化为一个多学科的研究课题，本文将该问题处理为确定从模态参数到模型参数之间的映射问题，并利用起源于地质统计学的Kriging技术实现该映射。对于一实际钢结构，通过实验测量得到了全部摸态数据，利用部分测量得到的模态数据进行了模型修改的实验研究，实验结果表明，利用修改后模型计算得到的全部摸态数据与测量结果吻合良好，表明该模型修改方法是可行的；同时该方法表现出了与基于神经网络的修改方法的互补特性。

Structural Model Updating Based on Kriging Methodology

Structural model updating has evolved as a multidisciplinary research subject. The problem is now treated as determining a mapping from modal parameters to structural model ones, and the mapping is realized by Kriging model originated in geological statistics. For a real steel structure, all modal data are measured experimentally, some model updating experiments are carried out with the aid of partially experimentally measured modal data. The experimental results indicate that modal data derived from updated model well coincide with the experimentally measured modal data, and show the complementary properties to neural network model updating method.