【作者】 袁旭东； 周晶；

【Author】 Yuan xu-dong , zhoujing (School of Civil & Hydraulic Engineering Dalian University of Technology, Dalian, 116024)

【机构】 大连理工大学水利土木学院；

【摘要】 为获得对连续结构损伤更加敏感并更好地应用于实际的损伤识别指标。在分析已有的识别指标基础上,选择基于结构模态参数的组合损伤识别指标作为神经网络输入参数,对结构的损伤位置、损伤程度进行了数值模拟研究。结果表明,利用该指标能对结构的损伤进行较为精确的识别和预测。更多还原

【Abstract】 In order to obtain more sensitive and more convenient signatures for structural damage detection, after the analysis of the current signatures for damage detection, the combined signatures of damage detection are selected as input parameters for neural networks. The numerical detection study for the damage location and extent of structures is carried out. It is proved that the structural damage is identified and predicted comparative accurately by this signature更多还原