Vi Congresso Nacional De Engenharia Mecânica Vi National Congress of Mechanical Engineering

Abstract. This paper deals with modal parameters identification using output-only data. A linear, time-invariant, finite dimensional mechanical system is considered, which is described by a stochastic state-space model excited by unknown operating forces. In this approach, the stochastic state-space model considers the errors due to state-variable and measurements, as integrant parts of the modeling, through a zero-mean white noise process. Subspace model identification consists in the approximation of a structured subspace described in terms of an extended observability matrix defined from a rank reduction process of a block Toeplitz matrix constructed from output-only data. This rank reduction procedure is calculated by using singular value decomposition. System matrices are evaluated using the shift-invariant property of the extended observability matrix. The natural frequencies, damping factors and modal shapes are identified by means of eigenvalues and eigenvectors of the state matrix, respectively. The performance of the presented approach is shown through simulation examples.