Stochastic Subspace Identification Method Using Fuzzy Clustering Techniques

In recent years, modal analysis has become a major technology in the quest for determination, improvement and optimization of dynamic characteristics of engineering structures. Operational Modal Analysis (OMA) is basically associated with the identification of modal parameters from the natural responses. In OMA the loads are unknown and the modal parameters have to be identified using responses only. Stochastic Subspace Identification (SSI) method is one of the known OMA techniques in time domain in which stabilization diagrams are used to specify the optimum model order of SSI method which eliminates spurious numerical poles. Therefore, only true physical system poles are left. Recently, a new stabilization diagram is proposed in which all the poles are plotted against damping ratio and the physical poles are estimated using Fuzzy Clustering Technique. In this paper, SSI method is applied to the numerical model of a structure and the stabilization diagram is plotted using the conventional approach. Then fuzzy clustering is used for plotting a new view of stabilization diagram. The obtained results from two methods are compared showing that the new method is more efficient and less time consuming.

کلیدواژه ها:

Fuzzy Clustering ، Stabilization Diagram ، Modal Analysis ، Stochastic Subspace IdentificationMethod