Structural damage detection method of CFRP plate based on SSA, LPP and K-Means algorithms

Abstract

The damage detection technology of Carbon Fiber Reinforced Polymer (CFRP) structure is of great significance for timely and accurate detection of structural damage and to ensure the safe service of the structure. In this paper, a structural damage detection method for CFRP based on SSA, LPP and K-Means was proposed. Firstly, the Lamb wave response signal of CFRP structure was analyzed, and a damage feature extraction method based on singular spectrum analysis (SSA) and locality preserving projection (LPP) was proposed. On this basis, a damage detection method based on K-Means clustering was proposed to cluster the normal state samples. Then the clustering distance boundary of the normal state samples was obtained to realize damage detection. Finally, a simulation case was designed to verify the proposed method. The results showed that for 150 groups of test samples, the method proposed in this paper was correct. This paper provided a feasible method for structural damage detection of CFRP plate.