摘要
对基于模态参数的结构计算模型修正算法而言;实验振型扩充是一个必要环节。由于修正问题需要迭代求解,所以以往算法用于振型扩充的系统矩阵一般都来自于上一迭代过程,导致系统矩阵误差传递给用作修正参考值的扩充实验振型,滞缓了迭代收敛速度。针对这一问题,提出一种改进算法,通过给扩充后的振型矩阵添加一个修正项来减小因系统矩阵不精确而导致的扩充振型误差,振型修正项与待修正参数之间的相关关系由求解一个最小化系统特征方程和振型误差的优化问题获得。算例研究表明,改进的算法可以提高修正的收敛速度,且迭代收敛曲线更趋平滑。
For modal parameter based on model updating algorithms ,experimental modeshape expansion is one of the indispensable steps. Since model updating generally means an iterative procedure, system matrices used in modeshape expansion always come from the previous iteration and consequently transfer errors to the expanded modeshapes. To reduce these errors, an improved updating procedure was proposed in this paper. A correction was added to the expanded modeshapes and the relation of this correction to the design parameters can be extracted by an optimism problem to minimize the deviation of modeshapes and system equations. Numerical example was introduced to verify the effectiveness and feasibility of the new updating procedure.
出处
《同济大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2002年第11期1295-1299,共5页
Journal of Tongji University:Natural Science
