Classification of roller defects and damages based on Rayleigh wave signals and power spectrum images with specific sampling rate

It is important to detect the defects and damages in the rollers and repair it since the defects and damages of rollers has a negative impact on the quality of the rolled products. Ultrasonic testing technology has the advantages of large detection depth, accurate defect localization, low cost, convenient use, fast speed, and harmless to the human body. This article proposes a deep learning classification method of defects and damages based on Rayleigh wave signals and power spectrum images with specific sampling rate, automatic identification of four common types of defects and damages is achieved by establishing end-to-end learning models. Firstly, an organic glass inclined block and clamping device were designed. In the experiment, time-domain signals were received on the right side of the defected or damaged samples, and signal data sets and power spectrum image data sets were established for signals with different sampling rates. Next, a defect and damage detection model is established based on a deep learning framework, which includes ResNet, GoogLeNet, DenseNet, and AlexNet with convolutional channels to extract signal features for classifying defects and damages. Finally, the performances of DenseNet models with different structures and depths are compared based on key indicators such as accuracy and training time. The experimental demonstrates that under higher sampling rate conditions, using the power spectrum images as data input yields better results than directly using Rayleigh wave signals. Moreover, for the power spectrum images of 5 M Rayleigh waves, using ResNet-18 to establish a deep learning model can achieve a classification accuracy of up to 0.998 and shorter training time.