论文部分内容阅读
利用小波分析与人工神经网络相结合的方法，对结构缺损进行了识别。敲击缺损试件后产生的振动信号由传感器拾取，经数据采集系统采集、适当处理后进行小波变换，形成人工神经网络的训练样本，并对所建网络进行训练，利用在训练样本中加入随机噪音的方法对网络的识别精度进行了讨论。本文用上述方法以悬臂梁试件为例，对结构缺损位置、缺损深度以及螺栓联接结构的紧固程度进行了试验。
Using the method of combining wavelet analysis and artificial neural network, the structural defect was identified. Vibration signals generated by percussive defect specimens are picked up by sensors and collected by data acquisition system. After appropriate processing, wavelet transform is performed to form training samples of artificial neural network, and the network is trained. By adding random The method of noise discusses the recognition accuracy of the network. In this paper, the cantilever specimen is taken as an example to test the position of the defect, the depth of the defect and the degree of fastening of the bolt connection structure.