节点文献
基于神经网络的桁架损伤识别
The Study of Truss Structure Damage Detection Using Neural Network
【摘要】 基于模态振型,提出一个用于结构损伤识别的参数-柔度差值曲率,并将之作为神经网络的输入向量。然后通过ANSYS建立了一个平面桁架结构的数学模型,用有限元分析计算结果作为神经网络的学习样本和测试样本,对不同程度的模拟损伤使用该神经网络进行识别。通过验证,表明该方法用于损伤识别简单有效。
【Abstract】 In this paper,based on mode shape,the author put forward a new parameter-flexibility difference curvature for detecting structure damages.And it is also the input vector of the neural network.Then a numerical model of truss structure based on ANSYS is built up,and through finite element analysis and calculating,we get the training samples and forecasting samples for the network.The neural network can identify the damage location and degree of the structure.It proves that this method is simple and valid.
【关键词】 神经网络;
柔度差值曲率;
损伤识别;
桁架;
【Key words】 neural network; flexibility difference curvature; damage detection; truss structure;
【Key words】 neural network; flexibility difference curvature; damage detection; truss structure;
【基金】 河北省博士基金项目(04547001D-5)
- 【文献出处】 潍坊学院学报 ,Journal of Weifang University , 编辑部邮箱 ,2010年04期
- 【分类号】TU323.4;TU312.3
- 【被引频次】1
- 【下载频次】51