| 识别混凝土板内部缺陷的人工神经网络算法 |
| |
| 引用本文： | 王婷,赵鸣,李杰. 识别混凝土板内部缺陷的人工神经网络算法[J]. 同济大学学报(自然科学版), 2007, 35(3): 304-308 |
| |
| 作者姓名： | 王婷 赵鸣 李杰 |
| |
| 作者单位： | 同济大学,建筑工程系,上海,200092 |
| |
| 基金项目： | 国家自然科学基金;土木工程防灾国家创新研究群体资助项目 |
| |
| 摘 要： | 以一维有缺陷混凝土板为研究对象，分别采用Leverberg-Marcluardt和径向基神经网络算法，对缺陷的深度与厚度进行识别，从而实现对混凝土板内部缺陷的三维重构，称为红外CT模拟．两类神经网络算法的识别结果表明：Leverberg-Marcluardt神经网络较径向基神经网络具有更好的收敛精度与计算效率．
|
| 关 键 词： | 结构检测 红外成像 红外CT模拟 神经网络 |
| 文章编号： | 0253-374X（2007）03-0304-05 |
| 修稿时间： | 2005-06-27 |
Identifying Interior Damage of Concrete Slabs Based on Artificial Neural Network |
| |
| WANG Ting,ZHAO Ming,LI Jie. Identifying Interior Damage of Concrete Slabs Based on Artificial Neural Network[J]. Journal of Tongji University(Natural Science), 2007, 35(3): 304-308 |
| |
| Authors: | WANG Ting ZHAO Ming LI Jie |
| |
| Affiliation: | Department of Building Engineering, Tongji University, Shanghai 200092,China |
| |
| Abstract: | Two kinds of algorithms, including Leverberg-Marquardt and Radial Basis Function neural network, are utilized to identify the depth and thickness of defect to realize three dimensional reconstruction of defects of concrete slabs, which is called infrared CT. In addition, a comparison of the two algorithms in identifying the depth and thickness of defect shows that. Leverberg-Marquardt neural network is precisie in convergence and efficient in computation. |
| |
| Keywords: | structural detection infrared thermography infrared CT simulation neural network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
