Online crack evaluation method based on guided wave time-frequency spectrum difference and convolutional neural network set

Abstract

The invention discloses an online crack evaluation method based on a guided wave time-frequency spectrum difference and a convolutional neural network set. The method comprises the following steps: acquiring a reference guided wave signal when no crack exists and monitoring guided wave signals under different crack lengths; calculating a time-frequency spectrum difference map of the reference guided wave signal and the monitoring guided wave signal as the input of the convolutional neural network to obtain a guided wave time-frequency spectrum difference map-crack length training data sample; obtaining a trained convolutional neural network set; and calculating a time-frequency spectrum difference map of the on-line monitoring guided wave signal and the on-line reference guided wave signal, and inputting the time-frequency spectrum difference map into the convolutional neural network set to realize on-line monitoring of the fatigue crack length of the structure. According to the method, a guided wave signal time-frequency spectrum difference map is obtained through a third-order complex Gaussian wavelet to serve as the input of the convolutional neural network; and forming a convolutional neural network set by adopting a plurality of convolutional neural networks, and taking average output of the plurality of networks as a structural fatigue crack diagnosis result.

Description

technical field [0001] The invention belongs to the technical field of structural health monitoring, and in particular relates to an online crack evaluation method based on guided wave time spectrum difference and convolutional neural network set. Background technique [0002] Metal structures have always been the main load-carrying structures of aviation vehicles, and their safety and reliability are of great strategic significance for national defense security and national economic development. For aerospace metal structures, fatigue cracks are one of the most important and dangerous damage forms leading to their failure. Under the action of alternating loads and corrosive environments, the structure may initiate fatigue cracks and gradually expand. Fatigue cracks appearing in key parts will seriously weaken the bearing capacity of the structure, and its instability and damage may even lead to catastrophic accidents. Therefore, timely and accurate online acquisition of s...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem of online monitoring of fatigue cracks in aerospace metal structures, the purpose of the present invention is to provide an online crack evaluation method based on the time-spectrum difference of guided waves and convolutional neural network sets, and obtain the time-spectrum of guided-wave signals through third-order complex Gaussian wavelets The difference map is used as the input of the convolutional neural network; multiple convolutional neural networks are used to form a convolutional neural network set, and the average output of multiple networks is used as the diagnosis result of structural fatigue cracks

Images