Crack Detection Method of Sleeper Based on Cascade Convolutional Neural Network

This work presents a new method for sleeper crack identification based on cascade convolutional neural network (CNN) to address the problem of low efficiency and poor accuracy in the traditional detection method of sleeper crack identification. The proposed algorithm mainly includes improved You Only Look Once version 3 (YOLOv3) and the crack recognition network, where the crack recognition network includes two modules, the crack encoder-decoder network (CEDNet) and the crack residual refinement network (CRRNet). The improved YOLOv3 network is used to identify and locate cracks on sleepers and segment them after the sleeper on the ballast bed is extracted by using the gray projection method. The sleeper is inputted into CEDNet for crack feature extraction to predict the coarse crack saliency map. The prediction graph is inputted into CRRNet to improve its edge information and local region to achieve optimization. The accuracy of the crack identification model is improved by using a mixed loss function of binary cross-entropy (BCE), structural similarity index measure (SSIM), and intersection over union (IOU). Results show that this method can accurately detect the sleeper crack image. During object detection, the proposed method is compared with YOLOv3 in terms of directly locating sleeper cracks. It has an accuracy of 96.3%, a recall rate of 91.2%, a mean average precision (mAP) of 91.5%, and frames per second (FPS) of 76.6/s. In the crack extraction part, the F-weighted is 0.831, mean absolute error (MAE) is 0.0157, and area under the curve (AUC) is 0.9453. The proposed method has better recognition, higher efficiency, and robustness compared with the other network models.

3D crack identification using the Nelder-Mead Simplex algorithm combined with a random generation of crack positions

In this paper, we present a scheme for cracks identification in three-dimensional linear elastic mechanical components. The scheme uses a boundary element method for solving the forward problem and the Nelder-Mead simplex numerical optimization algorithm coupled with a low discrepancy sequence in order to identify an embedded crack. The crack detection process is achieved through minimizing an objective function defined as the difference between measured strains and computed ones, at some specific sensors on the domain boundaries. Through the optimization procedure, the crack surface is modelled by geometrical parameters, which serve as identity variables. Numerical simulations are conducted to determine the identity parameters of an embedded elliptical crack, with measures randomly perturbed and the residual norm regularized in order to provide an efficient and numerically stable solution to measurement noise. The accuracy of this method is investigated in the identification of cracks over two examples. Through the treated examples, we showed that the method exhibits good stability with respect to measurement noise and convergent results could be achieved without restrictions on the selected initial values of the crack parameters.

Crack identification in beam by antiresonant frequencies

The present paper deals with the concept of antiresonance in multiple cracked beams and application for multi-crack identification. First, governing equations for antiresonant frequency are conducted and used for both computing antiresonant frequencies versus crack parameters and measuring-loading colocation and identifying cracks by measured antiresonant frequencies. Then, a procedure is proposed for crack identification in cantilever beam by antiresonant frequencies based on the so-called crack scanning method. Theoretical development is illustrated by numerical examples.

Frequency Contour-Strip Method for Characterization of Damage in Structures under Noisy Conditions

Eigen-frequency, compared with mode shape and damping, is a more practical and reliable dynamic feature to portray structural damage. The frequency contour-line method relying on this feature is a representative method to identify damage in beam-type structures. Although this method has been increasingly applied in the area of damage identification, it has two significant deficiencies: inefficiency in establishing the eigen-frequency panorama; and incompetence to identify cracks in noisy conditions, considerably impairing the effectiveness in identifying structural damage. To overcome these deficiencies, a novel method, termed the frequency contour-strip method, is developed for the first time. This method is derived by extending the frequency contour line of 1D to frequency contour strip of 2D. The advantages of the frequency contour-strip method are twofold: (i) it uses the isosurface function to instantly produce the eigen-frequency panorama with a computational efficiency several orders of magnitude higher than that of the frequency contour-line method; and (ii) it can accommodate the effect of random noise on damage identification, thereby thoroughly overcoming the deficiencies of the frequency contour-line method. With these merits, the frequency contour-strip method can characterize damage in beam-type structures with more efficiency, greater accuracy, and stronger robustness against noise. The proof of concept of the proposed method is performed on an analytical model of a Timoshenko beam bearing a crack and the effectiveness of the method is experimentally validated via crack identification in a steel beam.

Deep Transfer Learning Approach for Identifying Slope Surface Cracks

Geohazards such as landslides, which are often accompanied by surface cracks, have caused great harm to public safety and property. If these surface cracks could be identified in time, this would be of great significance for the monitoring and early warning of geohazards. Currently, the most common method for crack identification is manual detection, which has low efficiency and accuracy. In this paper, a deep transfer learning approach is proposed to effectively and efficiently identify slope surface cracks for the sake of fast monitoring and early warning of geohazards, such as landslides. The essential idea is to employ transfer learning by training (a) a large sample dataset of concrete cracks and (b) a small sample dataset of soil and rock masses’ cracks. In the proposed approach, (1) pretrained crack identification models are constructed based on a large sample dataset of concrete cracks; (2) refined crack identification models are further constructed based on a small sample dataset of soil and rock masses’ cracks. The proposed approach could be applied to conduct UAV surveys on high and steep slopes to provide monitoring and early warning of landslides to ensure the safety of people and property.