结构计算模型修正的区间反演方法

Interval Method for Computational Model Updating of Dynamic Structures

提出了结构计算模型修正的区间反演方法。该方法将计算模型修正问题归结为非线性全局优化问题 ,并采用一种遗传算法进行求解。方法考虑了作为修正参考值的试验模态参数具有一定误差 ,计算结果除可以给出修正后模型的估计值外还可以同时给出其不确定度 ,有利于对修正后计算模型的质量做出直观评价。算例表明本文提出的方法是可靠可行的 ,不仅可以判断采用的信息是否完备 。

By using the interval number to model a structure based on uncertain measured data, a new uncertain inverse method, interval back analysis method, is proposed for computational model updating of dynamic structures. The problem is turned into a nonlinear global optimization problem, and a hybrid genetic algorithm is introduced to solve the problem. By using the proposed method, not only the appropriate solutions but also the uncertainty degree of the solutions can be obtained. The quality assessment of the solutions is convenient to carry out within the method. The complements of the adopted data and the stability of the solutions can be assessed numerically.