首页 | 本学科首页   官方微博 | 高级检索  
     

梁损伤小波包分析和神经网络识别
引用本文:孟范孔,邱志成.梁损伤小波包分析和神经网络识别[J].噪声与振动控制,2013,33(1):197-200.
作者姓名:孟范孔  邱志成
作者单位:( 华南理工大学 机械与汽车工程学院,广州 510641 )
基金项目:国家自然科学基金资助项目(基金编号:60934001)
摘    要:针对柔性悬臂梁裂缝损伤问题进行损伤位置和损伤程度的识别研究。首先用有限元法建立系统动力学模型。然后对系统的动力响应信号进行小波包分解,建立基于小波包能量谱的损伤指标。把损伤指标作为改进BP神经网络的输入特征参数,用分步识别方法进行损伤位置和损伤程度的识别。最后进行了数值仿真研究。仿真结果表明,利用小波包分析和改进的BP神经网络可以精确地识别出柔性梁的损伤位置和损伤程度。

关 键 词:振动与波  损伤识别  小波包分析  改进的BP神经网络  有限元法  
收稿时间:2012-04-10

Damage Identification of Beam Using Wavelet Packet Analysis and Neural Network
MENG Fan-kong,QIU Zhi-cheng.Damage Identification of Beam Using Wavelet Packet Analysis and Neural Network[J].Noise and Vibration Control,2013,33(1):197-200.
Authors:MENG Fan-kong  QIU Zhi-cheng
Affiliation:(School of Mechanical and Automotive Engineering,South China University of Technology,Guangzhou 510641,China)
Abstract:A study on recognition of damage location and extent for cracked flexible cantilever beam was conducted. Firstly, finite element method was utilized to model the dynamic system. Then, the dynamic response signal was decomposed by using wavelet packet, and the damage indexes were calculated based on wavelet packet energy spectrum. These damage indexes were used as improved BPNN’s input characteristic parameters to recognize the damage location and extent by a two-step identification method. Finally, numerical simulation was carried out. The simulation results demonstrate that the damage location and damage extent of the flexible beam can be recognized accurately by using wavelet packet analysis and improved BP neural network.
Keywords:
本文献已被 CNKI 维普 等数据库收录!
点击此处可从《噪声与振动控制》浏览原始摘要信息
点击此处可从《噪声与振动控制》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号