Multi-decoder full convolutional neural network and a corresponding mesoscopic structure identification method thereof

Abstract

The invention provides a multi-decoder full convolutional neural network and a corresponding microscopic structure identification method thereof, the multi-decoder full convolutional neural network isused for semantic segmentation of ceramic matrix composite material perform XCT slices, and an initialization method, a class balance method, a training method and the like of a network structure aredetermined. According to the method, the problem of edge blur or content information loss in the feature map caused by different decoders is solved, the accuracy of microscomic structure recognitionis improved by combining deep learning and semantic segmentation tasks, and the robustness of an algorithm is enhanced. The multi-decoder full convolutional neural network provided by the invention can be used for ceramic matrix composite preforms of different weaving types, including 2.5-dimensional weaving, three-dimensional four-direction weaving structures and other all types, the applicationrange is wide, and identified microscopic structures comprise fiber bundles (warp yarns and weft yarns), holes and matrixes.

Description

technical field [0001] The invention belongs to the field of mesoscopic structure recognition of ceramic matrix composite material preforms, and in particular relates to a multi-decoder fully convolutional neural network for semantic segmentation of XCT slices of woven ceramic matrix composite materials. Background technique [0002] The structure of complex preforms of ceramic matrix composites (CMCs) includes 2.5D weaving, three-dimensional four-way weaving and other structures. The structure and the mesoscopic components of the preform determine the mechanical properties and failure mechanism of the material. In the finite element analysis model It has become a trend to consider the internal real structure of materials as accurately as possible. [0003] XCT scanning is a non-destructive testing method, which can accurately observe the real microstructure inside the material without destroying the material. A series of XCT slices of the scanned ceramic matrix composite m...

AI Technical Summary

Problems solved by technology

[0009] Aiming at the deficiencies in the prior art, the present invention provides a multi-decoder fully convolutional neural network for semantic segmentation of XCT slices of ceramic matrix composite material preforms, and determines the initialization method, category balance, training method, etc. of the network structure. , solve the problem of blurred edges or loss of content information in feature maps caused by different decoders, combine deep learning and semantic segmentation tasks to improve the accuracy of mesostructure recognition, and enhance the robustness of the algorithm

Images