论文部分内容阅读
提出了一种基于神经网络的结构动力模型修改和破损诊断方法，讨论并解决了该法实施中的若干技术问题，如神经网络的拓扑结构和快速算法，结构模态特性的量化比较，结构破损特征信息的提取和处理，计算结果精度的提高等等。数值算例表明：该法对于结构动力模型的修改以及结构破损部位和程度的诊断非常方便和有效
A new method of structural dynamic model modification and damage detection based on neural network is proposed. Several technical problems in the implementation of this method are discussed and solved, such as topology and fast algorithm of neural network, quantitative comparison of structural modal characteristics, structural damage Feature information extraction and processing, calculation accuracy and so on. Numerical examples show that this method is very convenient and effective for the modification of structural dynamic model and the diagnosis of the location and degree of structural damage