摘要
本文基于作者提出的含层间分层损伤层合板的动力有限元分析模型和方法 ,计算了分层长度和位置对含层间分层损伤层合板结构的固有频率的影响 ,然后应用MATLAB的神经网络工具箱建立了人工神经网络 ,通过典型结构的仿真结果比较 ,证明了采用有限元动力分析和BP网络技术相结合的方法是一种可用于复合材料层合板的分层损伤诊断的有效方法。
A strategy for detecting delamination of composite laminates is proposed in this paper, which is based on the dynamic finite element method in conjunction of neural networks scheme. Firstly the influences of the length and location of delamination upon the natural frequency of the composite laminates are calculated by a nonlinear dynamic finite element method for the delaminated composite laminates developed by authors. Then,the back propagation (BP) neural network in the Toolbox of MATLAB software is established for training and testing the network. From the simulating results it can be shown that the present strategy is efficient for detecting delamination of the composite laminates.
出处
《玻璃钢／复合材料》
CAS
CSCD
2004年第5期3-6,共4页
Fiber Reinforced Plastics/Composites
基金
国家自然科学基金项目 (1 0 2 72 0 2 5)
关键词
动力有限元分析
神经网络
复合材料层合板
损伤诊断
delamination detection
nonlinear dynamic finite element analysis
neural network scheme
composite laminates
