Open Access

ARTICLE

A set-based method for structural eigenvalue analysis using Kriging model and PSO algorithm

Zichun Yang^1,2,3, Wencai Sun²

1 Huazhong University of Science and Technology, Wuhan, 430074
2 Naval University of Engineering
3 Center for Aerospace and Education, University of California, Irvine

Computer Modeling in Engineering & Sciences 2013, 92(2), 193-212. https://doi.org/10.3970/cmes.2013.092.193

Abstract

The set-based structural eigenvalue problem is defined, by expressing the uncertainties of the structural parameters in terms of various convex sets. A new method based on Kriging model and Particle Swarm Optimization (PSO) is proposed for solving this problem. The introduction of the Kriging model into this approach can effectively reduce the computational burden especially for largescale structures. The solutions of the non-linear and non-monotonic problems are more accurate than those obtained by other methods in the literature with the PSO algorithm. The experimental points for Kriging model are sampled according to Latin hypercube sampling method. Two approaches of imposing the constraint of the hyper-ellipsoid are presented for global optimization. One is by adding penalty terms to the original objective function; the other one is use objective function with interval spherical coordinates by coordinate transformation. An engineering example revealed the feasibility and accuracy of the proposed method.

---

Keywords

---