Affiliation of Author(s):航空学院

Journal:J Vib Shock

Abstract:Here, the FE model updating method based on surrogate model was studied. An augmented support vector machine (SVM) based on hybrid basis functions was proposed for solving over-fitting results when SVM was used to deal with weak nonlinear functions. Based on dynamic test results measured and calculated results with a structural finite element model, according to design requirements, sensitivity analysis or engineering experience, appropriate parameters to be modified and modification ranges were chosen to determine the modification sample space and sample points. Then, the surrogate model for each group of sample points and corresponding objective function was constructed adopting the augmented SVM. The multi-objective optimization algorithm based on Pareto optimal solution was introduced to find the global optimal solution to parameters to be modified within the modification interval taking the output of the surrogate model as the objective and the sample space as variables. Example 1 showed that the prediction results with the augmented SVM have a higher accuracy than those with the traditional SVM do. Example 2 and 3 showed that the structural dynamic model updating based on the proposed augmented SVM is valuable in actual application and its results have a higher precision. © 2017, Editorial Office of Journal of Vibration and Shock. All right reserved.

ISSN No.:1000-3835

Translation or Not:no

Date of Publication:2017-08-15

Co-author:Huan He,cgp

Correspondence Author:Zhe Chen