Nhảy đến nội dung

An effective damage identification procedure using model updating technique and multi-objective optimization algorithm for structures made of functionally graded materials


D. Dinh-Cong, T. Nguyen-Thoi

Source title: 
Engineering with Computers, 2021 (ISI)
Academic year of acceptance: 

The application of multi-objective optimization algorithms in the field of structural damage identification has gained increasing attention in the past few years. Nevertheless, their application to the damage detection problems of composite structures is still very limited. In this regard, the article presents the first attempt to implement a multi-objective optimization framework based on multi-objective cuckoo search (MOCS) algorithm for identifying the locations and extent of multi-damages in structures made of functionally graded materials. First, we cast the structural damage identification procedure into an optimization-based FE model updating problem, where two sub-objective functions, namely, flexibility matrix change (Flex) and modal assurance criterion (MAC), are established for the multi-objective optimization purpose. Then, the MOCS as an effective optimizer is adopted to solve the multi-objective optimization, which results in a set of Pareto-optimal solutions for damage identification. Subsequently, a decision-making process is made for finding the most preferred solution in the Pareto-optimal set. Finally, numerical simulation studies on a two-span continuous FGM beam and a cantilever FGM plate are conducted to investigate the feasibility and accuracy of the proposed damage identification procedure. According to the obtained identification results, the proposed procedure can yield good predictions for the damage locations and corresponding severities in both single and multi-damage cases of the FGM structures using spatially incomplete measurement data with noise contamination. In addition, the results also show that the MOCS algorithm provides a better damage prediction than two other well-known algorithms, including Non-dominated Sorting Genetic Algorithm II (NSGA-II) and Multi-objective Grey Wolf Optimizer (MGWO).