K-means Optimizer: An Efficient Optimization Algorithm for Predicting the Uncertain Material Parameters in Real Structures

Abstract

AbstractIn this paper, an efficient optimization algorithm is used to predict the uncertain material parameters based on finite element (FE) model updating. For this purpose, an objective function is determined based on the difference between the dynamic characteristics of the measurement and FE results. Then, this objective function is optimized to determine the model’s uncertain material parameters. K-means optimizer (KO), a powerful algorithm developed by the authors in previous work, is used to handle this procedure. Then, the model's uncertain material parameters are determined by optimizing the objective function. The feature of KO is that it creates a strong balance between the abilities of exploration and exploitation. At the same time, this algorithm has also significantly improved the escape from local optima, which is appreciated to solve the complex problems. To evaluate the effectiveness of the proposed method, a practical example is used, i.e., a simple steel beam. The results obtained in this study have proved the effectiveness of the proposed method. At the same time, the KO algorithm is also shown to be a powerful algorithm for solving the optimization problems.KeywordsMetaheuristic optimizationK-means optimizerFE model updating