基于改进BPNN复合材料层合板损伤识别

Damage Identification of Composite Materials Based on Modified BP Neural Network

将B样条小波展开技术提取的损伤特征值作为改进BPNN输入进行学习与识别,识别结果显示该方法能够对复合材料层合板多种损伤进行快速准确的识别。

A dynamic approach based on modified BPNN for damage identification of composite materials is proposed. By using wavelet series, the features of signals are extracted and input to modified BPNN for training the network and identifying the damages. The results show that the method can exactly identify the faults.