STUDY ON MODEL ORDER DETERMINATION OF THIN PLATE SYSTEMS WITH PARAMETER UNCERTAINTIES

In this paper, the robustness analysis in determining the model order of thin plate systems with parameter uncertainties is presented in two ways. One is from the test data directly using Hankel singular values, and average component cost, and the other is from the uncertain model using the relative degree curlt and robustness index zeta , which are related to the singular values of controllability and observability grammians. ARMAX approach with real-time identification is adopted and modal tests are carried out to identify modal parameters for analysis. Experimental results show the effectiveness and robustness of the proposed methods. It provides a practical guidance to the modelling and dynamic analysis of thin-plate systems with parameter uncertainties.