Finite Element Model Updating Based on Least Squares Support Vector Machines

  • Authors:

  • Yue Zhu;Lingmi Zhang

  • Affiliations:

  • College of Aerospace Engineering, Nanjing University of Aeronautics & Astronautics, Nanjing,;College of Aerospace Engineering, Nanjing University of Aeronautics & Astronautics, Nanjing,

  • Venue:
  • ISNN 2009 Proceedings of the 6th International Symposium on Neural Networks: Advances in Neural Networks - Part II
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Finite element model updating based on the design parameter is a kind of inverse problem in structural dynamics, whose theoretical foundation is using the features of the structure to be a function of design parameters. According to the first-order derivative of the features with respect to design parameters, iterative solution is made. This paper presents a new method which treats the model updating as a positive problem. Features are independent variables and design parameters are dependent variables. The least squares support vector machines (LS-SVM) is utilized as a map function. The objective value of the design parameters can be directly estimated due to the generalization character of the LS-SVM. The method avoids solving the complicated nonlinear optimization problem which is difficult in the reported methods. Finite element model updating based on LS-SVM about the GARTEUR aircraft model is studied. Simulation results show the errors of design parameters and modal frequencies are less than 2% and 1%, respectively.