期刊文献+

基于曲率模态信息熵和BP神经网络的简支梁损伤识别方法 被引量:6

Damage identification method for simply supported beam based on curvature modal information entropy and BP neural network
下载PDF
导出
摘要 针对曲率模态指标存在的对振型节点附近损伤不敏感、需要较密集测点等缺陷,在曲率模态的基础上引入广义局部信息熵,推导了广义局部曲率模态信息熵的公式,建立了相应的损伤指标并针对高阶振型的特点对指标进行改进.提取并处理简支梁损伤模型的动力参数,将曲率模态和广义局部曲率模态信息熵分别作为神经网络的输入参数,对损伤进行识别并对比了两种参数的识别结果,研究了测点数量对指标精确度的影响,通过损伤钢梁模态试验对上述结论进行验证.结果表明,所提方法能较好地定位及定量损伤,在靠近振型节点处指标识别精度高于曲率模态. In view of the defects of curvature modal index,such as insensitivity to the damage near the vibration node and the necessity for intensive measuring points,the generalized local information entropy was introduced on the basis of curvature mode,the formula of generalized local curvature modal information entropy was deduced,and the corresponding damage index was established and improved according to the characteristics of high-order vibration mode.The dynamic parameters of damage model for the simply supported beam were extracted and processed.The curvature mode and the generalized local curvature modal information entropy were used as the input parameters of neural network to identify the damage,and the recognition results for two parameters were compared.The influence of the number of measuring points on the index accuracy was studied,and the above-mentioned results were verified by the damaged steel beam modal test.The results show that the as-proposed method can better locate and quantify the damage,and the index accuracy near the vibration nodes is higher than that of curvature mode.
作者 项长生 原子 周宇 李凌云 XIANG Chang-sheng;YUAN Zi;ZHOU Yu;LI Ling-yun(School of Civil Engineering,Lanzhou University of Technology,Lanzhou 730050,China;Western Civil Engineering Research Center of Disaster Prevention and Mitigation of the Ministry of Education,Lanzhou University of Technology,Lanzhou 730050,China;School of Civil Engineering,Anhui University of Architecture,Hefei 230601,China)
出处 《沈阳工业大学学报》 CAS 北大核心 2022年第5期575-583,共9页 Journal of Shenyang University of Technology
基金 国家自然科学基金项目(51868045) 安徽省高校省级自然科学研究重点项目(KJ2019A0746).
关键词 广义局部曲率模态 信息熵 曲率模态 神经网络 损伤定位 损伤定量 模态测试 简支梁 generalized local curvature mode information entropy curvature mode neural network damage location damage quantification modal test simply supported beam
  • 相关文献

参考文献7

二级参考文献41

共引文献237

同被引文献56

引证文献6

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部