Prestressed pipeline grouting quality detection method based on ultrasonic waves

The invention discloses a prestressed pipeline grouting quality detection method based on ultrasonic waves. The method is characterized in that pre-burying piezoelectric ceramic transducers inside and on the surface of a prestressed pipeline, and respectively using for transmitting and receiving ultrasonic signals; adopting wavelet packet transformation to decompose ultrasonic signals layer by layer, and constructing an energy distribution matrix of all wavelet packet components; and establishing a convolutional neural network model, and inputting an energy distribution matrix formed by wavelet packet components into the model to evaluate the grouting quality of the prestressed pipeline. According to the method, identification of the long-distance grouting defect type and the defect size in the length direction of the prestressed pipeline can be achieved. 本发明公开了一种基于超声波的预应力管道灌浆质量检测方法，其特征是将压电陶瓷换能器预埋在预应力管道内部和表面，分别用以发射和接收超声波信号；采用小波包变换逐层分解超声波信号，构建所有小波包分量的能量分布矩阵；建立卷积神经网络模型，将小波包分量组成的能量分布矩阵输入到模型中进行预应力管道灌浆质量的评估；本发明方法能