【毕业设计、论文】基于小波神经网络的锚杆锚固质量分析.pdfVIP

第 25 卷 第 1 期 岩石力学与工程学报 Vol.25 No.1 2006 年 1 月 Chinese Journal of Rock Mechanics and Engineering Jan .，2006 基于小波神经网络的锚杆锚固质量分析 刘明贵，岳向红 ( 中国科学院 武汉岩土力学研究所，湖北 武汉 430071) 摘要：敲击–回波法正广泛应用于锚杆锚固质量的无损检测中，该方法中存在诸多的影响因素，因此很难直接根 据实测的时程曲线对锚杆锚固质量进行准确的人工判读。首先，将小波分析和人工神经网络结合起来，采用数值 模拟得到锚杆外端部的动态响应并进行小波分析，从各分量中提取特征向量；然后，将这些特征向量输入人工神 经网络进行训练。分析结果表明，训练后的神经网络能根据实测信号的特征向量对锚杆锚固质量进行智能化分类。 关键词：数值分析；敲击–回波法；小波分析；神经网络；锚固质量 中图分类号：O 241 文献标识码：A 文章编号：1000 –6915(2006)01 –0083 –05 ANALYSIS OF BOLT′S ANCHORING QUALITY BASED ON WAVELET NEURAL NETWORK LIU Ming-gui ，YUE Xiang-hong (Institute of Rock and Soil Mechanics ，Chinese Academy of Sciences，Wuhan，Hubei 430071 ，China) Abstract ：The dynamic response of bolt achieved by impact-echo method can be effected easily by some factors. It′s difficult to classify bolt′s anchoring quality according to the dynamic response of bolt. In this paper，a new analytical method of bolt′s dynamic response is put forward. The new analytical method combing with advantages of wavelet analysis and artificial neural network is presented. Firstly ，the dynamic response of bolt by numerical 3D simulation software FLAC is obtained. Secondly ，the features are extracted from the dynamic response of bolt by wavelet analysis. Thirdly ，these features are fed into the artificial neural network for training. Finally，this artificial neural network to classify the bolt′s anchoring quality intelligent according to the dynamic response of bolt is employed. In order to validate this new analytical method，some practical signals of bolt′s dynamic response are used for the test. The results indicate that the new analytical method