利用试验模态分析及神经网络技术对结构损伤检测的探讨

选用一工程上常用的截面开口槽形梁，通过试验模态分析对结构损伤引起的模态频率移动进行了分析，从试验角度对四种损伤检测方法进行了有效性比较。并尝试将神经网络与试验模态分析技术相结合应用到结构损伤定量检测的研究中，不仅验证了其可行性，而且指出应变指标对结构损伤更敏感。

STUDY OF STRUCTURAL DAMAGE DETECTION USING EXPERIMENTAL MODAL ANALYSIS AND NEURAL NETWORKS TECHNOLOGY

Modal analysis technology draws more and more attention in the development of study in structural damage detection. In most cases, emphasis has been on using a finite element model of simple uniform cross sectioned beam and many means have been developed. In this paper an opening notch formed thin wall beam was chosen to provide data for analysis. The shift of modal frequency caused by structural damage was studied based on experimental modal analysis. In view of Experiment, the comparison of four damage detection means' effectiveness was presented. The combination of neural networks and modal analysis was introduced into the study on quantitative detection of structural damage. And the results reinforced the feasibility of the combination and indicated that the strain signature is more sensitive to damage.