A robust technique for damage identification of marine fiberglass rectangular composite plates using 2-D discrete wavelet transform and radial basis function networks. (1st November 2022)

• Record Type: Journal Article
• Authors: Saadatmorad, Morteza; Jafari-Talookolaei, Ramazan-Ali; Pashaei, Mohammad-Hadi; Khatir, Samir; Abdel Wahab, Magd
• Is Part Of: Ocean engineering. Volume 263(2022)
• Publication Date: 2022-11-01
• Subjects: Damage detection -- Wavelet transform -- Damaged laminated composite plates -- WT-RBFN -- Wavelet-artificial intelligence -- Radial basis function networks; Ocean engineering -- Periodicals; Ocean engineering; Periodicals; 620.4162
• DOI: 10.1016/j.oceaneng.2022.112317
• Languages: English
• ISSNs: 0029-8018

Abstract

Abstract: Although the classical wavelet technique can detect the location of damages with high accuracy in medium and high-level damages, their accuracy in damage detection decreases as the level of damages decreases. Thus, one of the significant challenges in damage detection by classical wavelet transform is the presence of noise in results. The presence of noise in wavelet transforms causes confusion in identifying the actual location of damaged structures. As a novel research, in order to eliminate this problem, the objective of this study is to propose a new and efficient technique called the wavelet transform-based radial basis function networks (WT-RBFNs) for robust damage detection of marine fiberglass rectangular laminated composite plates (RLCPs). A finite element model of single-damaged RLCPs is developed to generate single-damage vibration amplitude signals. The two-dimensional discrete wavelet transform is used to decompose these signals and detect the location of damages. By considering various single-damage scenarios, numerical and experimental findings show that the proposed WT-RBFN can efficiently eliminate the weakness of damage detection by the classical two-dimensional discrete wavelet transform, even for very low-level of damages. The proposed method can efficiently detect marine fiberglass RLCPs for practical damage detection. Highlights: New improved ANN based on WT. Damage identification in laminated composite. Robust and fast prediction of damage.