Damage Clustering Method of CFRP Structure Based on PCA and SOM Neural Network

Abstract

The damage identification of carbon fiber reinforced polymer (CFRP) structure is of great significance for guaranteeing the safety of the structure. In this paper, the damage clustering method of CFRP Structure based on principal component analysis (PCA) and self-organizing feature map (SOM) was studied. Firstly, the active excitation method was used to stimulate the CFRP structure and the structural response signals were obtained. Then, the structural response signals were transformed by Fourier transform and the principal component analysis was carried out to get the structural damage features. On this basis, the SOM neural network was used to cluster the damage features, and the CFRP structure damage clustering was realized. At last, the experimental system was constructed to verify the method proposed in this paper. The results showed that the damage clustering method based on PCA and SOM can accurately cluster the CFRP damage. This paper provides a feasible method for damage clustering of CFRP structure.