No abstract

The present study investigated ice accretion thickness under non-incoming flow icing conditions on the ground using an infrared thermography system that converts infrared radiation temperature. Two back propagation (BP) neural network models were developed to measure ice thickness. Both theoretical model and polynomials were employed to fit the icing surface temperature elevation sequence to extract the pixel-level temperature attenuation characteristics, which were served as the input to the BP neural networks. The prediction method of ice thickness by the BP neural network was analysed from three perspectives of sensitivity, dimension, and precision. In addition, K-nearest neighbour (KNN) and support vector regression (SVR) algorithms were compared with BP neural network. In terms of prediction effect, the BP neural network performed the best. The verification of the BP neural network based on the characteristics of the theoretical model proved that the method can effectively predict the thickness of ice accretion, and its prediction error does not exceed 10%.

Active infrared thermography has been extensively employed in non-destructive testing in a wide variety of fields. It is capable of extracting defect information of tested object based on the infrared thermal image sequence. However, conventional infrared thermal images are often subjected to defect information with low pixel resolution, and defects are difficult to quantitatively analyze. By exploiting flat-bottomed holes in a PVC plate as defect specimens, a method for quantitative defect depth recognition based on the fusion principal component analysis algorithm with sliding-window mechanism and the 1DResnet50-CBAM (One Dimensional - Residual Neural Network - Convolutional Block Attention Module) model was proposed for the reconstructed image sequence of active infrared thermography to address the above-described issues in this study. First, defect information and location were extracted from the original infrared sequence thermal image of the specimen using principal component analysis algorithm with sliding-window mechanism. Then, the dimension of the defect data was reduced using the temporal characteristic of the infrared temperature field. That is, the three-dimensional (3D) defect data were transformed into one-dimensional (1D) temporal infrared thermal signal. Moveover, the 1D infrared signal time series corresponding to the defect pixel points in the infrared sequence image served as the input to the network, and the defect depth served as the output for automatic defect detection and depth quantification. As indicated by the results, the proposed method based on the fusion principal component analysis algorithm with sliding-window mechanism and 1DResnet50-CBAM model is capable of accurately detecting and quantifying defects. Compared with conventional prediction algorithms, the proposed model can more effectively extract defect information from the infrared detection images, with the defect depth relative prediction error less than 1.5%. Thus, the proposed model was confirmed as an effective method and model for defect recognition and quantitative analysis using infrared thermal detection technology.

No abstract

No abstract